Fluoride removal using a chelating resin containing phosphonic-sulfonic acid bifunctional group

Journal Pre-proof

Fluoride Removal Using a Chelating Resin Containing Phosphonic-Sulfonic Acid Bifunctional Group Rong Li , Xunan Tian , Imtiaz Ashraf , Bin Chen PII: DOI: Reference:

S0021-9673(19)31126-4 https://doi.org/10.1016/j.chroma.2019.460697 CHROMA 460697

To appear in:

Journal of Chromatography A

Received date: Revised date: Accepted date:

23 August 2019 26 September 2019 8 November 2019

Please cite this article as: Rong Li , Xunan Tian , Imtiaz Ashraf , Bin Chen , Fluoride Removal Using a Chelating Resin Containing Phosphonic-Sulfonic Acid Bifunctional Group, Journal of Chromatography A (2019), doi: https://doi.org/10.1016/j.chroma.2019.460697

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V.

Highlights   

A novel chelating resin with bifunctional groups was applied for fluoride removal. Resin exhibited better adsorption to fluoride compared with commonly used resins. Adsorption mechanism of the present resin for fluoride was speculated.

1

Fluoride Removal Using a Chelating Resin Containing Phosphonic-Sulfonic Acid Bifunctional Group Rong Li, Xunan Tian, Imtiaz Ashraf, Bin Chen* School of Chemical Engineering, Northwest University, Xi’an, Shaanxi Province 710069, P. R. China *

Corresponding author information.

Professor Bin Chen Mailing address: School of Chemical Engineering, Northwest University, Taibai North Road 229, Xi’an 710069, P. R. China. Phone: +86-15349223972 E-mail: [email protected]

2

Abstract S9570-Fe(III), a modified chelating resin containing sulphonated monophosphonic acid bifunctional groups, was used for the fluoride removal from aqueous phase for the first time. The results specified that S9570-Fe(III) exhibited better adsorption towards the fluoride ions as compared to the other commonly used chelating resins having monofunctional group such as iminodiacetic acid, sulfonic acid or carboxylic acid. Adsorption thermodynamic and kinetic studies of S9570-Fe(III) chelating resin for the fluoride also have been carried out. The thermodynamic results demonstrated that the adsorption was a spontaneous process accompanied with a gradual decrease in entropy and the low temperature was favorable for the fluoride ion adsorption. The kinetic experiments showed that the resin exhibited a rapid initial adsorption behavior and the adsorption process more complied with the pseudo-second order reaction model which indicating that the whole adsorption process was controlled by a combined mechanism of intraparticle diffusion and chemical sorption. Adsorption mechanism of S9570-Fe(III) resin for fluoride ions was predicted. The study demonstrated the effectiveness of the phosphoric-sulfonic acid bifunctional group chelating resin to remove fluoride, and provided a novel type removal method for the fluoride. Keywords: fluoride, phosphoric-sulfonic acid bifunctional group, chelating resin, sorption behavior

1. Introduction Fluoride ion which is one of the essential trace elements present in the drinking water is more concerned for researchers as accumulation cause problem in human body. Excessive or insufficient of fluoride concentration in the human body can develop lesions. Skeletal fluorosis and dental problem will happen when the fluoride intake is excessive (>1.5 mg/L) [1], which will cause spontaneous fractures and affect the functions of cells, adrenal glands and gonads [2]. Insufficient intake of fluoride (<1.0 mg/L) can cause diseases such as dental

3

caries and osteoporosis in the human body [3]. The concentration ranges of fluoride ions are different in different regions. The WHO (World Health Organization) recommends the concentration of fluoride in drinking water is 0.5~1.5 mg/L for human body [4]. In China, the appropriate concentration range is only 1.0~1.5 mg/L [5]. Recently, one report has shown that China ranks among the top countries in the world where drinking water-type fluorosis is most widespread and serious [6]. Therefore, reducing the fluoride content in drinking water has become a crucial problem to be solved. Three main methods for fluorine removal from aqueous phase are used: precipitation separation, membrane separation and adsorption separation. The precipitation separation method mainly includes coagulation sedimentation [7] and electrocoagulation methods [8] which remove the fluoride ions (F-) mainly by forming precipitates with iron or aluminum under different conditions. Similar to the chemical process of the traditional precipitation method, fluidized bed reactor (FBR) system also has been used to treat wastewater containing fluorine [9]. Membrane separation technology mainly uses electrodialysis [10] and reverse osmosis [11] to remove F-. Different from the traditional membrane technology, an efficient extraction method based on a bubble membrane has been developed for the deep removal of fluorine from wet-process phosphoric acid [12]. The adsorptive separation method commonly includes bone charcoal, activated alumina [13], zeolite, activated carbon, carbon ceramic composite [14,15] and ion exchange resin methods [1,16-18], which removes F- mainly by ion exchange characteristics and physicochemical adsorption. Low cost, simple operation and high efficiency of this method is considered most appropriate and active for fluoride adsorption from drinking water. 4

According to SHAB (hard-soft-acid-base) principle, strongly hydrated fluoride anion exhibits strong affinity towards the hard acid polyvalent metal ions as Al(III), La(III), Fe(III), Ce(III), Zr(IV), Ti(IV), Mn(II, IV) etc. Due to the above characteristics of fluoride, metal chelate-type adsorption has received more attentions as an effective defluorination technology in recent years. Millar G J et al. firstly reported the modification of iminodiacetic acid ion exchange resin to aluminum type, which selectively removed the F- from solution [19]. Popat K M et al. modified the chelating resin that contains aminomethylphosphonic acid group with aluminum metal ion for the removal of F- which proved that the metal cation has a sufficient effect for the removal of F - in water [2]. Viswanathan N et al. made the metal chelating modification of sulfonate-containing Indion FR 10 resin to study the removal of fluoride ion [20], which demonstrated that the metal chelating resin has better removal effects for fluoride. Currently, research hotspots of metal chelating adsorbents mainly focus on improving the defluorination efficiency of chelating functional groups. Thus, it is very meaningful for developing a new and active metal chelating resin for the defluorination of water. In this work, a modified metal chelating resin containing phosphoric-sulfonic acid bifunctional groups was proposed for F- removal from water for the first time. The effectiveness of the resin was evaluated by comparing and screening the defluorination results with commonly used chelating resins. The initial concentration, adsorption temperature and time for F- removal by this chelating resin were investigated. Kinetics and thermodynamics studies revealed the removal mechanism of F-.

1. Experimental 5

1.1 Materials and reagents FPC11, FPC22, PC3500 and IRC748 resins were supplied by Rohm & Hass, S9570 was provided by Purolite. The resins were sufficiently washed to remove residual impurities, followed by transformation with a 0.05 mol/L FeCl3 solution. Dry resin can be obtained by drying the transformed wet resins at 75℃ for 6 hr. Sodium fluoride (NaF), sodium chloride (NaCl), ferric chloride (FeCl3), ferrous sulfide (FeS), trisodium citrate (Na3C6H5O7.2H2O), glacial acetic acid (CH3COOH), sodium acetate (CH3COONa) and sodium hydroxide (NaOH) were acquired from Chemical Reagent Factory, Xi’an. All the chemicals were analytical grade. 1.2 Preparation of the solutions 2.2.1. Preparation of the fluoride-containing simulated solutions 0.2210 g of NaF was taken in 100 mL with ultra-pure water. 1 g/L NaF solution was used as a stock solution. According to the certain dilution ratios, fluoride-containing simulated solutions (10-20mg/L) were prepared, respectively [21]. 2.2.2. Preparation of the fluoride-containing standard solution 0.8400 g of NaF was added in 100 mL with ultra-pure water. 2×10-1 mol/L solution was used as a stock solution. According to the certain dilution ratios, 2×10-2, 2×10-3, 2×10-4 and 2×10-5 mol/L of fluoride-containing standard solutions were prepared, respectively. 2.3. Preparation of the chelating resins in Fe3+ form FPC11, FPC22, PC3500, IRC748 and S9570 were modified with 0.05 mol/L FeCl3 solution (1:8, v:v) at a flow rate of 2 BV/h for 30 min to obtain Fe 3+-type chelating resins. Fe3+-type chelating resins were washed with ultra-pure water till the effluents reached neutral, 6

and the leakage of Fe3+ in the effluents could not be detected. The obtained resins (FPC11-Fe(III), FPC22-Fe(III), FPC3500-Fe(III), IRC748-Fe(III), S9570-Fe（III）) were dried at 75℃ for 36 h, then were used for the sorption studies. 2.4. Adsorption experiments of the Fe3+-type chelating resins for F50 mL of fluoride-containing simulated solutions (15 mg/L) were added to 0.5 g of FPC11-Fe(III), FPC22-Fe(III), FPC3500-Fe(III), IRC748-Fe(III) and S9570-Fe(III) resins, respectively. The solutions were shaken in a water bath at 30°C for 24 h until the adsorption equilibriums were reached. The equilibrium concentrations of F - were measured, and the adsorption capacities of the resins were calculated according to Eq. (1). qe 

(C0  Ce )  V m

(1)

Where, qe is the equilibrium adsorption capacity of the chelating resin for F - (mg/gresin), Co is the initial concentrations (mg/L), Ce is the equilibrium concentrations (mg/L), V is the volume of the fluoride-containing simulated solution (L), and m is the weight of the chelating resin (g). 2.5. Adsorption experiment of S9570-Fe(III) resin for F2.5.1. Adsorption kinetics experiment 0.50 g of S9570-Fe(III) resins were added to 50 mL of the simulated solutions with different concentrations (10, 15 and 20 mg/L), respectively. Then kept in water bath at different temperature (303 - 333 K) separately and shaken. Samples were taken out after regular interval to check out the adsorption. Sample concentrations were calculated by the method given in 2.3 section, the adsorption capacities of the chelating resins for F- were calculated by Eq. (1). 7

2.5.2. Adsorption kinetics models 2.5.2.1. Pseudo-first order reaction model The model was used to describe the adsorption kinetics of the S9570-Fe(III) resin for F-. By substituting the boundary conditions (t=0, qt=0 and t=t, qt=qt) and integrating for the model [22], the linear Eq. (2) was obtained.

ln( qe  qt )  ln qe  k1 t

(2)

k1 is the rate constant (1/min), t is the adsorption time (min), qt and qe are the adsorption capacities at time t and equilibrium, respectively (mg/g). 2.5.2.2. Pseudo-second order reaction model For this model, we used the below equation which may help us to understand the adsorption of the resin for F- [23].

t 1 t   2 qt k2 qe qe

(3)

h  k2 qe 2

(4)

k2 is the rate constant (gresin/mg·min), h is the initial rate (mg/gresin·min). Based on the slope and intercept of the Eq. (3), qe and k2 can be obtained, respectively. 2.5.2.3. Elovich model The given equation is the simple form of Elovich model [24]. qt 

1



ln( ) 

1



ln t

(5)

α is the initial adsorption rate constant (gresin/mg·min2), β is the desorption rate constant (gresin/mg·min), which can be obtained by the intercept and slope of the Eq., respectively. 2.5.2.4. Intraparticle diffusion model 8

The intraparticle diffusion can be calculated by Weber-Morris equation [25-26]. qt  ki t1/2  C

(6)

ki is the intraparticle diffusion constant (mg/gresin min1/2). C is the characteristic parameter of intraparticle diffusion, and the value reflects the essence of adsorption process [25].

Ri  1 

C qe

(7)

Initial adsorption factor (Ri) can be calculated with the help of Eq. (7) [27]. 2.6. Adsorption thermodynamics of S9570-Fe(III) resin for F2.6.1. Adsorption thermodynamics experiment S9570-Fe(III) resin (0.1-3.0 g) was added into 15 mg/L fluoride ion solutions, respectively. The solutions were shaken at the speed of 200 rpm in a water bath at different temperature (303-333 K) for 15 h until adsorption equilibrium. The concentrations of F- were determined by the method described in 2.3 section, adsorption capacities of the resin for Fwere calculated by Eq. (1). 2.6.2. Adsorption isotherm models 2.6.2.1. Langmuir isotherm model Following assumptions are supposed for this model: the active sites on the surface of the adsorbent are homogeneous. Adsorption is a single layer adsorption, which only exists the interaction between the adsorbate and the active site of the adsorbent. The linear expression of the model is as follows [28-29]. Ce C 1   e qe qm K L qm

RL 

1 1  K LC0

(8)

(9) 9

qm is the adsorption capacity for single layer (mg/gresin), KL is the Langmuir isotherm constant (L/mg). The values of qm and KL can be determined from the slope and intercept of the linear plot of Ce/qe against Ce. RL is the characteristic separation coefficient. 2.6.2.2. Freundlich isotherm model This model assumes that the surface adsorption of the adsorbent is not uniform. The linear equation for this model is given as follows [30-31]: ln qe 

1 ln Ce  ln K F n

(10)

n is the characteristic constant, and KF is the Freundlich adsorption constant (mg/g resin). The values of n and KF were determined from the slope and intercept of the linear plot of ln qe against ln Ce. 2.6.2.3. Temkin isotherm model Assuming that the adsorption potential energy of the adsorbent surface is inconsistent, that is, the surface is not uniform. The model reflects the interaction among the adsorbates. The basic expression of the model is as follows [32]: qe  (

When

BT  (

RT ) bT ,

RT ) ln Ce KT bT

(11)

the linear expression of the model can be as follows:

qe  BT ln Ce  BT ln KT

(12)

bT is the adsorption heat for Temkin isotherm (J/mol), BT is the Temkin constant, KT is the equilibrium binding constant (gresin/mg). 2.6.2.4. Calculation of the thermodynamic parameter K0 

qe Ce

(13)

10

ln K0  

H S  RT R

G  H  T S

(14) (15)

Thermodynamics parameters (K0, ΔH, ΔS and ΔG) can be premeditated from above equations [33].

3. Results and discussion 3.1. Adsorption performance of the metal chelating resins for FIn order to investigate the factors affecting the removal of F- in the aqueous phase, we selected the different types of cation exchange resins to prepare the metal chelating resins. At present, the modification of a cation exchange resin with metal ions is mainly based on aluminum type [2, 19-20]. However, adsorption conditions of F- in the actual application process will cause leakage of aluminum ions, which can cause the complexity of the subsequent process. The sanitary standard for drinking water in China stipulates that the content of aluminum does not exceed 0.2 mg/L [5]. Considering the health for the human body, we selected and prepared the Fe3+-type chelating resins instead of AI3+-type, and compared the adsorption effects of Fe3+-type dry and wet resins for F- in the aqueous phase. As shown in Table 1, the matrixes of the resins were same, exchange capacities were similar. Therefore, pore size and functional group of the resins were the main investigation factors. From Fig. 1, the adsorption effect of all dry resins was significantly better than that of wet resins, indicating that dry resin favored for the adsorption of F -. Compared with that of FPC11-Fe(III), the adsorption capacity of FPC22-Fe(III) for F- was slightly lower, indicating that the effect of the pore size on the adsorption strength was not significant. The adsorption order of dry or wet resins for F- was consistent: S9570-Fe(III) > FPC11-Fe(III) > 11

IRC748-Fe(III) > FPC22-Fe(III) > FPC3500-Fe(III), which indicated that the type of functional group was the main reason for the difference in adsorption. Compared with the commonly used resins with monofunctional groups such as iminodiacetic acid (IRC748-Fe(III)), sulfonic acid (FPC11-Fe(III), FPC22-Fe(III)) and carboxylic acid (FPC3500-Fe(III)), S9570-Fe(III) with the phosphoric-sulfonic acid difunctional group exhibited superior removal effect for F-. It is indicated that the presence of multifunctional groups was more conducive to the removal of F - by the resin. Therefore, S9570-Fe(III) is worthy of further study to understand the adsorption mechanism of this difunctional group resin for F-. 3.2. Adsorption performance of S9570-Fe(III) for FGenerally, the initial concentration, temperature, and pH of the fluoride-containing liquid, particle size of the resin and stirring strength will affect the removal effect of F- by S9570-Fe(III) resin. Among these factors, the initial concentration and temperature are the main factors affecting the adsorption mechanism between the adsorbate and adsorbent. Therefore, under the premise of optimization of other conditions (pH: 5.20, particle size: 0.12
the equilibrium time shorted. 3.2.1. Adsorption kinetics of S9570-Fe(III) resin for FIn order to study the adsorption mechanism of S9570-Fe(III) resin for F- in aqueous phase, we used four kinetic models to interpret the adsorption behaviors of the resin for F-. From Table 2, compared with the pseudo-first order reaction and Elovich models, pseudo-second order reaction model had the highest fitting degree, r22 was above 0.99, the calculated values (qe,cal) and experimental values (qe,exp) of the model parameters were nearly consistent, indicating that the pseudo-second order reaction model could better describe the adsorption process of F- on S9570-Fe(III) resin. As we known, the pseudo-second order model assumes that there exists the electron-sharing or transfer chemical adsorption between the adsorbate and adsorbent. Therefore, it was speculated that there might exist coordination interaction or electrostatic interaction between the adsorbent and adsorbate [34]. In addition, ri12 of intraparticle diffusion model were above 0.93, indicating that the adsorption of F - on S9570-Fe(III) resin was a mixed mechanism controlled by both chemical adsorption and intraparticle diffusion [25, 34]. From the kinetic data of the Weber-Morris intraparticle diffusion equation, the equation displayed two-order linear characteristics, and ki1 > ki2, indicating that the adsorption process involved in multiple stages [35]. In the short time after the start of adsorption, the adsorption amount of S9570-Fe(III) resin for F- increased rapidly. The first-order fitting line of qt-t1/2 did not pass through the origin, indicating that F- was rapidly adsorbed onto the outer surface of the chelating resin particles at the beginning of adsorption, and intraparticle diffusion was not the only control step for the metal chelating resin to adsorb F- [36]. When the surface of the 13

resin reached saturation, F- gradually diffused into the pores of the chelating resin and was adsorbed to the active center of the resin, the adsorption amount of F- was gradually increased. ki1 > ki2 indicated that the adsorption of the resin for F- was slower and the equilibrium time was longer. Therefore, the adsorption process was controlled by the intraparticle diffusion. Commonly, if Ri =1, no rapid initial adsorption; 0.9< Ri <1, weak initial adsorption; 0.5< Ri <0.9, intermediate initial adsorption; 0.1< Ri <0.5, strong initial adsorption; Ri < 0.1 indicates that the adsorption process is completed in a short time [27]. From Table 3, 0.86 < Ri < 1 indicated that there was weak or intermediate initial adsorption. Through the above analyses, pseudo-second order reaction model and intraparticle diffusion model both displayed better suitability, which could be used to describe the adsorption behavior of S9570-Fe(III) resin for F- in the aqueous phase. 3.2.1.1. Effect of the initial fluoride concentration on the adsorption process of S9570-Fe(III) resin for FFrom Fig.3, Table 2 and Table 3, when the initial fluoride concentration increased from 10 to 20 mg/L, the adsorption driving force enhanced, h increased from 0.02 to 0.06 mg/g min, the equilibrium adsorption capacity increased, and k2 decreased. As shown in the fitting results of the pseudo-second order adsorption model, r22 were all greater than 0.99. From the fitting results of the intraparticle diffusion model, qt-t1/2 exhibited two-order linearity. With increasing fluoride concentrations, ki1 and ki2 increased, and Ri decreased, indicating that the initial adsorption of F- onto the outer surface of S9570-Fe(III) resin changed from weak initial adsorption to intermediate initial adsorption [25]. 3.2.1.2. Effect of the temperature on the adsorption process of S9570-Fe(III) resin for F14

As shown in Fig. 4, Table 2 and Table 3, when the adsorption temperature increased from 303 to 333 K, the equilibrium time was shortened and the amount of adsorption decreased which indicating that low temperature favored the adsorption of S9570-Fe(III) resin for F-. From the model fitting parameters, the temperature increased, k2 and h slightly increased, and r22 gradually increased, indicating that the dominant role of chemical adsorption control was enhanced [34]. Ri was between 0.86 and 0.96, indicating that S9570-Fe(III) resin displayed the initial adsorption strength for F- from intermediate to weak in the range of 303-333 K [27]. 3.2.2. Adsorption thermodynamics of S9570-Fe(III) resin for F3.2.2.1. Adsorption isotherms of S9570-Fe(III) resin for FFig.5 illustrates the fitting results of adsorption isotherm models for F- on S9570-Fe(III) resin at various temperatures. The relevant parameters are shown in Table 4. As shown in Fig.5 and Table 4, although the fitting results of the Freundlich model were the best (r2>0.99), the fitting effects of three adsorption isotherm models were all better, and r2 were greater than 0.97. As the temperature increased, qm, KL and KF all decreased in turn, indicating that the relative adsorption of S9570-Fe(III) resin for F- decreased with increasing temperature, that is low temperature favored adsorption. 0
3.2.2.2. Adsorption thermodynamic parameters of S9570-Fe(III) resin for FFrom Fig. 6, the linear relationship of qe against Ce was good. Thermodynamic parameters were calculated according to the method described in 2.7.2.4., and the results are shown in Table 5. As ΔH<0 indicated that the adsorption was an exothermic process, and the low temperature was favorable for the adsorption of F- onto the chelating resin. ΔG<0, signifying that the adsorption behavior of the resin for F- happened spontaneously. The absolute value of ΔG decreased as the temperature increased, demonstrating that the low temperature was favorable for the adsorption. ΔS<0, signifying that F- was adsorbed from the solution onto the resin surface, the degree of freedom reduced, and the chaos of the liquid-solid interface reduced. ΔH<0, ΔG<0 and ΔS<0, indicating that the adsorption process of the resin for F- was mainly influenced by enthalpy drive. 3.2.3. Adsorption mechanism of S9570-Fe(III) resin for FFrom the conclusions of the pseudo-second order reaction model as well as the functional group structure of the resin, we speculated that there might exist the coordination and electrostatic adsorptions between the resin and F-. As illustrated in Fig. 7, S9570 is a bifunctional cation exchange resin containing both phosphoric and sulfonic acid groups, which were negatively charged and contained coordinating atoms to provide lone pair electrons. Since Fe3+ was positively charged and could provide empty orbits, both phosphoric and sulfonic acid groups could form the metal chelates with Fe3+ possibly by the coordination or electrostatic interaction, or maybe both of them. On the other hand, F- could provide a lone pair electron and was negatively charged. 16

Utilizing the remaining orbital or positive charge of Fe3+, it was possible further to adsorb Fby the coordination or electrostatic action, thereby removing F - from the aqueous phase. Compared with traditional monofunctional resins such as sulfonic acid-containing, phosphoric acid-containing and iminodiacetic acid-containing resins (Fig. 1), the bifunctional resin with both monophosphonic and sulfonic acid functional groups performed better for Fremoval from the aqueous solution. This is possibly because sulphonated monophosphonic resin structure. The hydrophilic sulphonic acid group together with non-hydrophilic monophosphonic acid group in polymer structure had a dual-mechanism in sorption, where the first group increased the kinetics of sorption, where the other group selectively chose the iron ion [37-38]. As a result, S9570 resin could chelate more iron ions, thereby adsorbing more fluoride ions, indicating that the presence of multifunctional groups was more conducive to the removal of F-. 4. Conclusions Compared with the commonly used resins with monofunctional groups, S9570-Fe(III) resin with both monophosphonic-sulfonic acid bifunctional groups performed best for the fluoride removal from the aqueous solution. Kinetic study demonstrated that the whole adsorption process was a combination mechanism of chemical sorption and intraparticle diffusion. Thermodynamic study proposed that the adsorption was an exothermic process, low temperature was promising for the adsorption of F- on the chelating resin, and the adsorption was a spontaneous process companied with a decreasing entropy gradually.

17

Due to the sulphonated monophosphonic structure of S9570-Fe(III) resin, the adsorption mechanism of F- onto the chelating resin involved the coordination or electrostatic interaction, or maybe both of them.

Notes The authors declare no competing financial interest.

Declaration of interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding: This work was supported by the National Natural Science Foundation of China (No.21376191); the Service Local Special Project of Shaanxi Province Education Department (No.14JF027); the Key Research and Development Program of Shaanxi Provincial Science and Technology Department (No.2018GY-079); and the Industrialization Cultivation Item of Shaanxi Province Educational Department (N0.2013jc23).

References [1] M.T. Samadi, M. Zarrabi, M.N. Sepehr, S.M. Ramhormozi, S. Azizian, A. Amrane, Removal of fluoride ions by ion exchange resin: kinetic and equilibrium studies, Environ. Eng. Manage. J. 13 (2013) 205-214. [2] K.M. Popat, P.S. Anand, B.D. Dasare, Selective removal of fluoride ions from water by the aluminium form of the aminomethylphosphonic acid-type ion exchanger, React. Polym. 23 (1994) 23-32. [3] W. Yang, The present situation and prospect of drinking water defluorinated technology, 18

Tech. Develop. Enter. 31 (2012) 63-87. [4] WHO (World Health Organization), Guidelines for drinking water quality health criteria and supporting information, second ed., Geneva, Switzerland, 1996. [5] GB5749–85, Sanitary standard for drinking water, Beijing: China Standard Press, 2006. [6] Y. Gao, Current difficulties in prevention and control of drinking water type of endemic fluorosis and countermeasures, Chin. J. Endem. 37 (2018) 4-6. [7] M. Wang, The progress of aluminum salt, Jiangxi Chem. Industry, 2 (2005) 9-11. [8] L. Sun, X. Wan, X. Chen, Thesis and research defluorination by electrocoagulationtheory and practice, Chi. Wat. Wastewater. 2 (1988) 5-9. [9] G. Zeng, B. Ling, Z. Li, S. Luo, X. Sui, Q. Guan, Fluorine removal and calcium fluoride recovery from rare-earth smelting wastewater using fluidized bed crystallization process, J. Hazard. Mater. 373 (2019) 313-320. [10] S.K. Adhikary, U.K. Tipnis, W.P. Harkare, K.P. Govindan, Defluoridation during desalination of brackish water by electro dialysis, Desalination, 71 (1989) 301-302. [11] S.V. Joshi, S.H. Mehta, A.P. Rao, A.V. Rao, Estimation of sodium fluoride using HPLC in reverse osmosis experiments, Wat. Treat. 7 (1992) 207-211. [12] Y. Zuo, Q. Chen, C. Li, C. Kang, X. Lei, Removal of fluorine from wet-process phosphoric acid using a solvent extraction technique with tributyl phosphate and silicon oil, ACS OMEGA. 4 (2019) 11593-11601. [13] S. Meenakshi, Studies on defluoridation of water with a few adsorbents and development of an indigenous defluoridation unit for domestic use, Congress of Transportation & Development Institute, 2 (1992) 1224-1231. 19

[14] C.S. Sundaram, N. Viswanathan, S. Meenakshi, Uptake of fluoride by nanohydroxyapatite/chitosan, a bioinorganic composite, Bioresour. Technol. 99 (2008) 8226-8230. [15] Y.H. Li, S. Wang, X. Zhang, J. Wei, C. Xu, Z. Luan, D. Wu, Adsorption of fluoride from water by aligned carbon nanotubes, Mater. Res. Bull. 38 (2003) 469-476. [16] C. Montero-Ocampo, J.F.M. Villafañe, Effect of dissolved species on the fluoride electro-removal from groundwater, ECS Trans. 28 (2010) 57-65. [17] Y. Ku, H.M. Chiou, W. Wang, The removal of fluoride ion from aqueous solution by a cation synthetic resin, Sep. Sci. Technol. 37 (2002) 89-103. [18] I.B. Solangi, S. Memon, M.I. Bhanger, Removal of fluoride from aqueous environment by modified Amberlite resin, J. Hazard. Mater. 171 (2009) 815-819. [19] G.J. Millar, S.J. Couperthwaite, D.B. Wellner, D.C. Macfarlane, S.A. Dalzell, Removal of fluoride ions from solution by chelating resin with imino-diacetate functionality, J. Water Process Eng. 20 (2017) 113-122. [20] N. Viswanathan, S. Meenakshi, Role of metal ion incorporation in ion exchange resin on the selectivity of fluoride, J. Hazard. Mater. 162 (2009) 920-930. [21] Y. Jin, C.Y. EX, L. Zhang, C. Chen, Standard examination methods for drinking water-Nonmetal parameters (GB/T5750.5–2006), Beijing: China Standard Press, 2006. [22] M. Aliabadi, I. Khazaei, H. Fakhraee, M.T.H. Mousavian, Hexavalent chromium removal from aqueous solutions by using low-cost biological wastes: equilibrium and kinetic studies, Int. J. Environ. Sci. Technol. 9 (2012) 319-326. [23] G. Darracq, J. Baron, M. Joyeux, Kinetic and isotherm studies on perchlorate sorption by 20

ion-exchange resins in drinking water treatment, J. Water Process Eng. 3 (2014), 123–131. [24] C.W. Cheung, J.F. Porter, G. Mckay, Sorption kinetics for the removal of copper and zinc from effluents using bone char, Sep. Purif. Technol. 19 (2000) 55-64. [25]W.J. Weber Jr, J. Morris, Kinetics of adsorption on carbon from solutions, J. Sanitary Eng. Div, Proc. Am. Soc. Civil Eng. 89 (1963) 31-39. [26] C. Sarıcı-Özdemir, Y. Önal, Equilibrium, kinetic and thermodynamic adsorptions of the environmental pollutant tannic acid onto activated carbon, Desalination, 251 (2010) 146-152. [27] F.C. Wu, R.L. Tseng, R.S. Juang, Initial behavior of intraparticle diffusion model used in the description of adsorption kinetics, Chem. Eng. J. 153 (2009) 1-8. [28] A.H. Chen, Y.Y. Huang, Adsorption of Remazol Black 5 from aqueous solution by the templated crosslinked-chitosans, J. Hazard. Mater. 177 (2010) 668-675. [29] I. Langumuir, The adsorption of gases on plane surfaces of glass, mica and platinum, J. Am. Chem. Soc. 40 (1918) 1361-1403. [30] H.M.F. Freundlich, Over the adsorption in solution, J. Phys. Chem. 57 (1906) 358-471. [31] S. Vasiliu, I. Bunia, S. Racovita, V. Neagu, Adsorption of cefotaxime sodium salt on polymer coated ion exchange resin microparticles: Kinetics, equilibrium and thermodynamic studies, Carbohydr. Polym. 85 (2011) 376-387. [32] M.I. Temkin, Kinetics of ammonia synthesis on promoted iron catalysts, Acta physiochim. 12 (1940) 327-356. [33] G.L. Maddikeri, A.B. Pandit, P.R. Gogate, Adsorptive removal of saturated and 21

unsaturated fatty acids using ion-exchange resins, Ind. Eng. Chem. Res. 51 (2012) 6869–6876. [34] Y.S. Ho, Review of second-order models for adsorption systems, J. Hazard. Mater. 136 (2006) 681-689. [35] E. Lorenc-Grabowska, G. Gryglewicz, Adsorption of lignite-derived humic acids on coal-based mesoporous activated carbons, J. Colloid Interface Sci. 284 (2005) 416-423. [36] D. Mohan, K.P. Singh, S. Sinha, D. Gosh, Removal of pyridine from aqueous solution using low cost activated carbons derived from agricultural waste materials, Carbon. 42 (2004) 2409-2421. [37] R. Chiariza, E.P. Horwitz, S.D. Alexandrators, M.J. Gula, Diphonix resin: a review of its properties and applications, Sep. Sci. Technol. 32 (1997) 1-35. [38] A. Izadi, A. Mohebbi, M. Amiri, N. Izadi, Removal of iron ions from industrial copper raffinate and electrowinning electrolyte solutions by chemical precipitation and ion exchange, Miner. Eng. 113 (2017) 23-35.

22

Figure captions

Fig.1 Adsorption of F- by the metal chelating resins.

Fig.2 Effects of the initial concentration, temperature on F- adsorption onto S9570-Fe(III) resin.

23

Fig.3 Kinetic models of F- adsorption onto S9570-Fe(III) resin at various initial concentrations.

Fig.4 Kinetic models of F- adsorption onto S9570-Fe(III) resin at various temperatures.

24

Fig.5 Adsorption isotherm models of S9570-Fe(III) resin for F- at various temperatures.

Fig.6 Adsorption isotherm of S9570-Fe(III) resin for F-.

25

Fig.7 Adsorption schematic of S9570-Fe(III) resin for F-.

26

Table 1 Physical and chemical characteristics of the cation exchange resins. Resin FPC 11

Matrix polystyrene

Functional group

Pore size

Exchange capacity(eq/L)

-SO3H

gel

≥ 2.05

FPC 22

polystyrene

-SO3H

macroporous

≥ 1.80

FPC 3500

polystyrene

-COOH

macroporous

≥ 2.60

IRC 748

polystyrene

CH2COOH-NH-CH2COOH

macroporous

≥ 1.35

S9570

polystyrene

-SO3H & -PO3H2

macroporous

≥ 1.02

27

Table 2 Adsorption kinetics model parameters of S9570-Fe(III) resin for F-. pseudo-first order model

pseudo-second order model

-3

qe, cal

k2

h

r2 2

0.8583

0.79

0.08

0.05

0.9996

1.94

0.8363

0.73

0.09

0.05

0.9998

0.53

2.58

0.8631

0.71

0.12

0.06

0.9999

0.65

0.56

3.07

0.9297

0.66

0.12

0.06

0.9999

10

0.52

0.56

2.11

0.9651

0.53

0.09

0.02

0.9994

15

0.74

0.62

2.09

0.9586

0.74

0.07

0.04

0.9996

20

0.93

0.70

2.64

0.9594

0.93

0.07

0.06

0.9996

qe,exp

qe,cal

k1×10

r1

303

0.79

0.59

1.64

313

0.72

0.56

323

0.71

333

2

T(K)

C0(mg/L)

Elovich model

intraparticle diffusion model

α

β

r2

ki1×10-2

ki2×10-3

ri12

303

0.44

10.13

0.9382

7.36

4.23

0.9329

313

0.26

10.00

0.9075

8.15

2.63

0.9402

323

0.36

10.63

0.8957

7.75

1.62

0.9347

333

0.21

10.80

0.9162

7.17

2.26

0.9264

10

0.16

14.21

0.9698

6.36

5.40

0.9620

15

0.34

10.57

0.9635

9.09

7.28

0.9692

20

0.74

9.05

0.9561

10.73

7.78

0.9705

T(K)

C0(mg/L)

Table 3 Adsorption diffusion model parameters of S9570-Fe(III) resin for F-. ki1×10-2

Ci1

Ri

Initial adsorption behavior

303

7.36

0.11

0.86

intermediate initial adsorption

313

8.15

0.02

0.97

weak initial adsorption

323

7.75

0.06

0.92

weak initial adsorption

333

7.17

0.03

0.96

weak initial adsorption

10

6.36

0.002

0.996

weak initial adsorption

15

9.09

0.026

0.965

weak initial adsorption

20

10.73

0.093

0.900

intermediate initial adsorption

T(K)

C0(mg/L)

28

Table 4 Isotherm adsorption parameters of S9570-Fe(III) resin for F- at various temperatures. Model

T /K

Langmuir

Freundlich

Temkin

Regression equation

Parameter RL

KL

qm

r2

303

y=1.59300x+2.33465

0.09

0.6823

0.628

0.9701

313

y=1.69825x+2.54426

0.09

0.6675

0.589

0.9783

323

y=1.78569x+3.06178

0.10

0.5832

0.560

0.9898

333

y=1.85996x+3.27294

0.10

0.5683

0.538

0.9931

KF

1/n

r2

303

y=0.41193x-1.34795

0.2598

0.4119

0.9960

313

y=0.40138x-1.41099

0.2439

0.4014

0.9954

323

y=0.40466x-1.51393

0.2200

0.4047

0.9987

333

y=0.39086x-1.54999

0.2123

0.3909

0.9994

KT

BT

r2

303

y=0.13966x+0.26163

6.5100

0.1397

0.9699

313

y=0.13064x+0.24295

6.4218

0.1306

0.9714

323

y=0.12673x+0.21222

5.3366

0.1267

0.9884

333

y=0.12079x+0.20145

5.3003

0.1208

0.9929

Table 5 Adsorption thermodynamic parameters of S9570-Fe(III) resin for F-. T (K)

△H (kJ/mol)

△G (kJ/mol)

△S (J/mol·K)

303

-12.10

-10.67

-5.05

313

-12.10

-10.66

-5.05

323

-12.10

-10.44

-5.05

333

-12.10

-10.37

-5.05

29