Green Synthesis of Iron Nano-Impregnated Adsorbent for Fast Removal of Fluoride from Water

Journal of Molecular Liquids 211 (2015) 457–465

Contents lists available at ScienceDirect

Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq

Green Synthesis of Iron Nano-Impregnated Adsorbent for Fast Removal of Fluoride from Water Imran Ali a,⁎, Zeid A. ALOthman b, Mohd Marsin Sanagi c,d a

Department of Chemistry, Jamia Millia Islamia (Central University), New Delhi 110025, India Department of Chemistry, College of Science, King Saud University, Riyadh 11451, Kingdom of Saudi Arabia Separation Science and Technology Group (SepSTec), Department of Chemistry, Faculty of Science, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia d Ibnu Sina Institute for Fundamental Science Studies, Nanotechnology Research Alliance, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia b c

a r t i c l e

i n f o

Article history: Received 27 June 2015 Received in revised form 10 July 2015 Accepted 12 July 2015 Available online xxxx Keywords: Removal of fluoride Adsorption models Water treatment Iron nano-impregnated adsorbent Mechanism of adsorption

a b s t r a c t Iron nano-impregnated adsorbent was synthesized, characterized and applied for fluoride subtraction from the water. Maximum fluoride removal (90%) was at 4.0 mg/L concentration, 25.0 min. contact time, 7.0 pH, 2.5 g/L dose and 293 K temperature. Iron nanocomposite adsorbent was selective for fluoride removal. The experimental data obeyed Langmuir, Freundlich and Temkin models. The values ΔG° were −1.89, −0.86 and −0.74 kJ/mol at 293, 298 and 303 K temperatures. ΔH° value was −7.61 kJ/mol; indicating exothermic adsorption. ΔS° value was −2.30 x 10−2 kJ/mol K; a signal of entropy decrease during adsorption. The adsorption process was in the order of 293 N 298 N 303 K. Kinetic modeling confirmed pseudo-first-order and liquid film diffusion mechanisms. The mechanism of adsorption is also determined. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Fluoride is one of the most harmful pollutants in the ground water. Water at some places in the world is not fit for drinking purpose due to fluoride contaminant. Geological and various anthropogenic activities are responsible for the ground water contamination by fluoride. Industries discharging fluoride are super phosphate, zinc smelters, ceramic works, aluminum smelters, brickworks, coal-fired power plants, steel mills, uranium enrichment facilities and oil refineries [1]. As per WHO, the permissible limit of fluoride is 1.0 mg/L [2]. High concentration of fluoride causes poor development of infant’s brain, osteosclerosis, dental fluorosis, cancer and impairment in human beings [3]. Fluoride pollution is a global problem in various countries such as USA, Canada, Brazil, Pakistan, India, Sri Lanka, China, Thailand, Japan, New Zealand, and some countries of Africa and Europe continents. The fluoride problem is recognized globally, and Fig. 1 shows dental and skeletal fluorosis world wide. In view of these facts, it is essential to develop, fast, selective, economic and eco-friendly method for the removal of fluoride from water. There are some techniques employed for fluoride removal, but adsorption is considered as the best one due to its unique features [4–18]. Literature survey indicates few papers on fluoride removal by adsorption [19–25]. These methods have certain drawbacks such as poor adsorption capacities, long contact time, high dose and extremely

⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (I. Ali).

http://dx.doi.org/10.1016/j.molliq.2015.07.034 0167-7322/© 2015 Elsevier B.V. All rights reserved.

low or high pHs. These limitations could not make these methods feasible to remove fluoride in real life problems. Therefore, the attempts were made to develop iron nano-impregnated adsorbent by green technology. The developed adsorbent was used for the removal of fluoride from water. The results of this study are discussed herein. 2. Experimental 2.1. Chemicals, Reagents, and Instruments Sodium fluoride was obtained from Merck, Darmstadt, Germany. 1-Butyl-3-methylimidazolidium bromide was purchased from SigmaAldrich Co., USA. Deionized water was prepared using Millipore-Q, Bedford, MA, USA system. pH meter of Control Dynamics (Model APX175 E/C) was used to measure pH of the solutions. The centrifuge of Remi (model C-30BL) was used to separate the adsorbent. Powdered X-ray diffraction was carried out on Philips PX-1830 diffractometer using Cu Kα radiation (λ = 1.54 Å), a Cu filter on secondary optics, 25 kV voltage, 30 mA current and a proportional counter detector. The residual concentration of fluoride was determined by UV–vis. spectrophotometer (T80, P.G. Instrument Ltd., U.K.) at 570 nm wavelength as per standard procedure [26]. 2.2. Preparation of Iron Nano-impregnated Adsorbent Green technology was exploited to prepare iron nanoparticles (NPs) as per the standard procedure [27,28]. The black tea (100.0 g/L) was

458

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

Fig. 1. The dental and skeletal fluorosis globally.

heated at 80 °C for 1 h. The extract was filtered followed by addition of 0.20 M ferrous sulphate solution in 1:2 ratio. The solution was kept for 24 h. The NPs formed were separated, washed with deionized water three times and drying in an oven at 250 °C for 24 h. 500 mg of 1Butyl-3-methylimidazolidium bromide was dissolved in 100 mL acetate buffer (0.05 M, pH 4.5). 5.0 g NPs were transferred to 100 mL solutions of 1-butyl-3-methylimidazolidium bromide. It was sonicated for 24 hrs. The treated iron NPs were separated and washed with deionized water three times and dried in an oven at about 100 °C for 24 h. The prepared iron nano-impregnated particles were ready for use in adsorption experiments.

2.5. Adsorption Studies All the adsorption experiments were carried out on thermostatic water bath shaker at a fixed temperature for a given period. After adsorption, the solid and liquid parts were separated by centrifugation. Fluoride concentrations in the solution samples were determined by UV Visb. spectroscopy. Adsorption isotherms were studied in the range of 0.5-7.0 mg/L as a concentration with 1.0-10.0 pH range, 2.550 min. contact time, 0.5-5.0 g/L dose and 293–303 K temperatures. The different mathematical models were used to ascertain isothermal and kinetic parameters. The data obtained in batch studies was used to calculate the equilibrium fluoride uptake capacity. It was calculated using the following equation.

2.3. Characterization of Iron nano-impregnated Adsorbent The synthesized iron nano-impregnated particles were characterized by UV–vis. spectrometry, field emission scanning electron microscope (FESEM) and XRD techniques. The morphology of the material was ascertained by scanning electron microscope (FESEM). Images of samples were recorded at different magnifications at 10 kV operation. X-Ray diffraction (XRD) patterns of native and iron nano-impregnated particles were obtained using Philips PX-1830 diffractometer using Cu Kα radiation (λ = 1.54 Å), a Cu filter on secondary optics, 25 kV voltage and 30 mA current and a proportional counter detector. Iron nanoimpregnated particles were scanned from 10° to 80° 2θ at a scanning rate of 3° 2θ per minute.

Q e ¼ ðC0 −Ce Þ=m

ð1Þ

where, Q e is the amount (mg/g) of fluoride adsorbed at equilibrium. Co is initial concentration (mg/L). Ct is the equilibrium concentration (mg/L) at time ‘t’. m is the weight of adsorbent in g/L. The percentage removal of fluoride was calculated using the following equation.

% Removal ¼ ½ðC0 −Ce Þ= C0 100

ð2Þ

where, C0 and Ct have the usual meanings. 2.6. Kinetics Studies

2.4. Preparation of Fluoride Solutions The standard solution of fluoride (10.0 mg/L) was prepared in deionised water. Further dilution for UV–vis. spectrometry (0.505.0 mg/L) and adsorption studies (0.5-7.0 mg/L) were carried out in deionised water.

Kinetics for fluoride adsorption was analyzed by uptake of fluoride from aqueous solutions at different times. For studying adsorption isotherms, the various solutions of fluoride were stirred with a known amount of adsorbent till equilibrium. The residual fluoride concentration was determined by UV–vis. spectrometry. The batch tests were

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

459

carried out to compare the adsorptive capacities at different time intervals. A series of Erlenmeyer flasks (100 mL) having a definite volume of fluoride solution of known concentrations; were kept on a thermostatic water bath shaker. The known amount of nano-impregnated iron adsorbent was added to each flask with shaking. The solutions were centrifuged, and the supernatants were analyzed for fluoride. The blank runs with iron nano-impregnated adsorbent were carried out under identical experimental for the experimental errors.

3. Results and Discussion 3.1. Characterization of Iron nano-impregnated Adsorbent The formation of iron nanoparticles was ascertained by observing changes in the peak areas of tea polyphenols and caffeine at 205 and 275 nm, respectively. The intensities of these peaks were decreased by the addition of ferrous sulphate solution. This confirmed the formation of iron NPs. The FESEM images of iron NPs indicated spherical shape with a diameter ranging 40–50 nm (Fig. 2). The polyphenols and caffeine present in tea extract were responsible in the formation of iron NPs. It was due to their reducing and capping properties [27,28]. In XRD spectra, the peaks appeared at 25, and 29° θ related to FeOOH [iron(III) oxide-hydroxide] and Fe2O3 (maghemite) (Fig. 3). It is appealing to observe that the peaks disappeared after impregnating iron NPs with 1-butyl-3-methylimidazolidium bromide. It was because 1-butyl3-methylimidazolidium bromide reacted with iron NPs completely. This observation established the formation of the iron nanoimpregnated material.

3.2. Effect of concentration The concentration was optimized in the range of 0.5-7.0 mg/L. The other variables were contact time 25 min., pH 7.0, dose 2.5 g/L and temperature 293 K. The residual concentration of fluoride was measured as discussed above. This effect is shown in Fig. 4a, which shows maximum uptake of fluoride at 4.0 mg/L. It is apparent from this plot that adsorption increased rapidly from 0.5 to 4.0 mg/L concentration. The adsorption capacities were 0.2, 0.4, 0.6, 0.76, 0.94, 1.12, 1.30 and 1.44 mg/g at 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5 and 4.0 mg/L concentrations, respectively. The adsorption became constant after 4.0 mg/L concentration. There was no increase in adsorption capacities at high concentration (4.57.0 mg/L). Therefore, it was concluded that 4.0 mg/L was the optimized concentration; with 90% fluoride removal.

Fig. 3. XRD patterns of (a): native and (b): iron nano-impregnated particles.

3.3. Effect of contact time The effect of contact time was optimized by doing experiments from 2.5-50.0 minutes. The other experimental conditions were 4.0 mg/L concentration of fluoride, 2.5 g/L dose, 7.0 pH and 293 K temperature. The results of this effect are plotted in Fig. 4b. It is obvious from this figure the adsorption capacities were 0.7, 0.9, 1.10, 1.13, 1.40 and 1.44 mg/g at 2.5, 5, 10, 15, 20 and 25 minutes of contact time, respectively. Further increase in contact time could not augment more adsorption. The percentage removal at 25 minutes observed was 90%. For this reason, optimized contact time was 25 minutes. 3.4. Effect of pH pH of adsorption was optimized by doing experiments at different pHs (1–10). The other experimental variables were 4.0 mg/L concentration of fluoride, 2.5 g/L dose, 25 min. contact time and 293 K. The results of this effect are shown in Fig. 4c. This figure clearly indicates that the adsorption capacities were 0.2, 0.39, 0.59, 0.80, 1.08, 1.22 and 1.44 at pH 1, 2, 3, 4, 5, 6, and 7, respectively. Further increase in pH did not enhance the adsorption capacity. The percentage removal at pH 7.0 was 90%. Therefore, this pH was chosen as the best one. It was accomplished that the method is eco-friendly as most of the water bodies have pH 7.0. 3.5. Effect of dosage This effect was studied by varying dose from 0.5-5.0 g/L. The other variables were kept constant such as pH 7.0, 30.0 min. contact time, 4.0 mg/L concentration of fluoride at 293 K temperature. The results of this effect are plotted in Fig. 4d. This figure clearly shows adsorption capacities 0.3, 0.6, 0.98, 1.25 and 1.44 mg/g at 0.5, 1.0, 1.5, 2.0 and 2.5 g/L dose, respectively. Further increase in dose amount had not resulted in any boost in adsorption. The percentage removal was 90% at 2.5 g/L dose. Therefore, this dose was chosen as the optimized. Of course, this effect dictated that adsorption method was economic due to low adsorbent dose. Therefore, this method may be transferred economically at the industrial level. 3.6. Effect of temperature

Fig. 2. FESEM image of iron nano-impregnated particles.

The effect of temperature on the adsorption of fluoride was studied by taking 293, 298 and 303 K temperatures. The other experimental conditions were 4.0 mg/L fluoride, pH 7.0, dose 2.5 g/L and contact time 25 min. The outcomes of this effect are given in Fig. 4e. The

460

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

Amount Adsorbed (mg/g)

Amount Adsorbed (mg/g)

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

0

0 0

2

4 Initial Con. (mg/L)

6

8

0

10

20 30 40 Contact Time (min.)

(a)

60

(b)

1.6

1.6 Amount Adsorbed (mg/g)

Amount Adsorbed (mg/g)

50

1.4 1.2 1 0.8 0.6 0.4 0.2 0

1.4 1.2 1 0.8 0.6 0.4 0.2 0

1

2

3

4

5

6

7

8

9

10

0.5

1

pH

(c)

1.5

2

2.5 3 3.5 Dose (g/L)

4

4.5

5

(d)

Amount Adsorbed (mg/g)

1.6 1.4 1.2 1 0.8

293 K

0.6

298 K 303 K

0.4 0.2 0 0

2

4 6 Contact Time (min.)

8

(e) Fig. 4. Optimization of adsorption parameters of Fluoride Herbicide; (a): initial conc, (b): contact time, (c): pH, (d): dose and (e): temperature.

sorption of fluoride was decreased with increasing temperature; confirming exothermic adsorption process. Fluoride removal followed 293 N 298 N 303 K order. Of course, adsorption method was considered eco-friendly as most of water bodies have the temperature range of 293–303 K. 3.7. Effect of interfering ions Fluoride adsorption was carried out in Millipore water. But real life groundwater has many other identities such as sodium, magnesium, potassium, sulfate, nitrate, calcium, chloride, phosphate, etc. Hence, the effects of these ions on the adsorption of fluoride were considered. The experiments were also conceded in groundwater (laboratory tap

water). Groundwater quality of tap water was monitored prior to fluoride removal. Conductivity and pH of the ground water were 1.48 mS/cm and 7.19, respectively. Total hardness, alkalinity, sodium, potassium, calcium, magnesium and total dissolved solids (TDS) were 448.0, 280.0, 2.15, 1.84, 204.0, 244.0 mg/L, and 542.25, respectively. A decrease in the adsorption of fluoride was observed (2.0-3.5; Table 1). This decrease in fluoride adsorption might be owing to the competitive adsorption capacities between fluoride and interfering ions. 3.8. Adsorption Isotherms Models The results obtained from adsorption experiments were treated by well-known models. The different isotherms applied were Langmuir,

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

461

1=Q t ¼ 1=Xm bCt þ 1=Xm

Table 1 Effect of interfering ions on the uptake of fluoride. S. No.

Interfering ions

Maximum percent reduction in the uptake of

1. 2. 3. 4. 5. 6. 7. 8.

Sodium Potassium Calcium Magnesium Chloride Sulfate Nitrate Phosphate

2.1 2.0 1.6 1.1 3.5 3.0 2.9 2.0

where, Ct and Qt have usual meaning. Xm is a constant and expresses maximum monolayer adsorption capacity of adsorbent (mg/g). b (L/mg) is another constant related to binding energy of fluoride onto the active sites of adsorbent. The values of Xm give an idea of the active sites while b driving force at equilibrium. The values of Xm and b were calculated from slope and intercept of a straight line of the graph (1/Qt vs 1/Ct). Langmuir graphs of fluoride adsorption at 293, 298 and 303 K are given in Fig. 5a. Langmuir isotherm fits the adsorption data of fluoride onto the adsorbent at all the temperatures. The values of regression constants (R2) were 0.990, 0.886 and 0.884 at 293, 298 and 303 K, respectively. The values of Langmuir’s constant (b) were 5.29, 10.75 and 12.21 L/mg at 293, 298 and 303 K (Table 2). These values showed good binding capacities of fluoride at all the temperatures. The values of Xm were 2.18, 1.42 and 1.34 mg/g at these temperatures. The values of the dimensionless constant [separation factor (RL)] were calculated by the following equation.

Freundlich and Temkin. The analyses results obtained from these models are discussed in the following sub-sections.

3.8.1. Langmuir Model This model ascertains the relationship among the concentrations of surface adsorbed species to the number of active sites at equilibrium. Langmuir isotherm assumes adsorbate adsorbed at a fixed number of well-defined sites. No further adsorption is possible after a molecule of adsorbate occupies a site on the adsorbent. Additionally, all sites are equivalent by energy point of views; with no interaction among adsorbate molecules. The suitability of this model confirms monolayer adsorption on a homogeneous surface without interactions among the adsorbed species. This adsorption isotherm is given by the following equation.

RL ¼ 1=ð1 þ bCe Þ

ð4Þ

These values were 0.110, 0.065 and 0.060 at 293, 298 and 303 K, respectively. Lower values of RL than 1.0 show energetically favorable adsorption. 3.8.2. Freundlich Isotherm This isotherm is applicable for large numbers of adsorption sites with no constraint of monolayer only; with non-uniform distribution

1.6

0.25

1.4

293 K

1.2 1

298 K logQt

0.8 0.6

0.2

293 K

0.15

298 K

0.1

303 K

303 K

0.05 0

0.4

-1.5

0.2

-1

-0.5

0

-0.05

0

0.5

-0.1

0

5

10

15

1/Ct

logCt

(a)

-0.15

(b) 1.6 1.4 1.2 1 Qt

1/Qt

ð3Þ

0.8 293 K

0.6

298 K

0.4

303 K

0.2 0

-2.5

-2

-1.5

-1

-0.5

0

lnCt

(c) Fig. 5. The plots showing (a): Langmuir, (b): Freundlich and (c): Temkin for the removal of Fluoride Herbicide.

462

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

Table 2 Isotherm parameters values for fluoride. Temps.

20 °C 25 °C 30 °C

Langmuir isotherm

Freundlich isotherm

Temkin isotherm

Xm (mg/g)

b (L/mg)

RL

R2

kF (mg/g)

n (g/L)

R2

KT (L/g)

BT(kJ/mol)

R2

2.18 1.42 1.34

5.29 10.75 12.21

0.067 0.035 0.019

0.990 0.886 0.884

1.70 1.42 1.30

3.42 3.72 4.28

0.788 0.891 0.910

60.83 92.76 135.64

5.54 7.76 9.53

0.943 0.934 0.895

of adsorption heat and affinities over the heterogeneous surface. The amount of the adsorbate on the surface of the adsorbent is the abridgment of all sites. The adsorption energies decrease exponentially on the completion of adsorption phenomenon. This model is given by the following equation.

logQ e ¼ ð1=nÞ loge þ logk F

ð5Þ

where, kF [(mg/g)] and n are empirical constants (Freundlich constants). kF relates to the relative adsorption capacities of the adsorbent. n is adsorption intensity. The values of n from 1 to 10 show favorable adsorption. The plots of logQe vs logCe were straight lines and the intercept related to kF. Adsorption intensity was computed by slope (1/n). Freundlich isotherm plots for fluoride adsorption are shown in Fig. 5b; with constants values given in Table 2. The kF values were 1.70, 1.42 and 1.30 at 293, 298 and 303 K, respectively. The values of n were 3.42, 3.72 and 4.28; indicating favorable adsorption. The regression coefficients at all three temperatures were close to one, showing the best fitting of this model. 3.8.3. Temkin Isotherm This model describes adsorbate and adsorbent interactions. The adsorption heat of all the molecules in the layer decreases linearly with coverage owing to adsorbate-adsorbate interactions. Additionally, adsorption is characterized by a uniform distribution of binding energies; up to some maximum binding energy. This model is given by the following equation.

Q e ¼ ðRT=BT ÞlnCe þ ðRT=BT ÞlnKT

ð6Þ

where, KT (L/g) is the equilibrium binding constant presenting maximum binding energy. BT constant is related to the heat of adsorption. R is the ideal gas constant (0.008314 kJ/mol/ K). T is the temperature in Kelvin. A graph of Qt vs logCe for fluoride was linear (Fig. 5c). The constants BT and KT were calculated from the slope and the intercept, respectively (Table 2). The values of KT were 60.83, 92.76 and 135.64 L/g at 293, 298 and 303 K, respectively; confirming strong interactions between adsorbate and adsorbent. The values of BT were 5.54, 7.76 and 9.35; showing a small variation in the heat of adsorption. The values of the regression coefficient (R2) were close to one confirming the suitability of this model. 3.9. Thermodynamic study The energy, enthalpy and entropy are the indications of a thermodynamic process. Gibbs free energy change of adsorption is correlated to the equilibrium constant by Vant Hoff’s equation as given below.

equilibrium constant, respectively. By replacing K via Xm (Langmuir constant), the equation becomes as follows. ΔG0 ¼ −RT ln Xm

ð8Þ

The values of ΔG0 at 293, 298 and 303 K were calculated and observed as − 1.89, − 0.86 and − 0.74 kJmol−1 (Table 3). The negative values of ΔG0 established adsorption favorable and spontaneous. Gibb’s free energy of adsorption (ΔG0) is related with changes in entropy (ΔS0) and enthalpy (ΔH0) by the following equations. ΔG0 ¼ ΔH0 − TΔS0

ð9Þ

The above equation can be changed to the following form. ln ðQ  Þ ¼ ΔS0 =R −ΔH0 =RT

ð10Þ

A plot of lnQ° vs 1/T was a straight line (graph not shown). The intercept gave value of ΔS0 while slope corresponded to ΔH0. The values of ΔH0 and ΔS0 were −7.61 and −2.30x10−2, respectively. The negative value of enthalpy change showed exothermic adsorption. The low negative value of ΔS0 showed decreased entropy of adsorption process. Thus, it was understood that fluoride adsorption was connected with the decline in mobility freedom of fluoride. 3.10. Kinetic Modeling The adsorption mechanism is established by kinetic modeling. It depends on the physical and chemical features of the adsorbate and adsorbent. Thus, various models were tested using adsorption data. These are discussed in the following sub-sections. 3.10.1. Pseudo-First-Order Kinetic Model This is the first model used for the mechanism determination. It is given by the following equation. dQ t =dt ¼ k1 ðQ e −Q t Þ

ð11Þ

The above equation was integrated with boundary conditions, t = 0 with Qt = 0 and t = t with Qt = Qt, to obtain lineralized equation as given below.

ln ðQ e −Q t Þ ¼ ln e –k1 t

ð12Þ

where, Qe, Qt and k1 (min−1) are the solute amount (mg/g) adsorbed at equilibrium, at time any time and equilibrium rate constant. These values were calculated and given in Table 4. The rate constant was

Table 3 Thermodynamic parameters values for fluoride.

ΔG0 ¼ −RT ln K

ð7Þ

where, ΔG0, T, R, and K are free energy change (kJ/mol), absolute temperature (K), universal gas constant (0.008314 kJ mol−1 K−1) and

ΔG° (kJ/mol) T = 293 K

T = 298 K

T = 303 K

−1.89

−0.86

−0.74

ΔH° (kJ/mol)

ΔS° (kJ/mol K)

−7.61

−2.30 x 10−2

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465 Table 4 Kinetic parameters for fluoride adsorption. Kinetic model

Kinetic parameters

Numerical value

Pseudo-first-order kinetic model

k1 (min−1) Experimental Qe (mg/g) Theoretical Qe (mg/g) R2 k2 (gmg−1 min−1) Experimental Qe (mg/g) Theoretical Qe (mg/g) h (mgg−1 min−1) R2 α (mgg−1 min−1) β (gmg−1) R2 kipd1 (mgg−1 min−0.5) Intercept R2 kfd (gmg−1) Intercept R2

0.07 1.44 1.66 0.931 0.016 1.44 2.33 0.087 0.965 5.07 0.52 0.916 0.28 0.266 0.991 0.071 0.14 0.939

Pseudo-second-order kinetic model

Elovich kinetic model

Intraparticle diffusion kinetic model Film diffusion kinetic model

3.10.2. Pseudo-Second Order Kinetic Model The attempts were also made to test pseudo-second order kinetic model for the experimental data. The pseudo-second order kinetic model for the adsorption phenomenon is given by the following equation.

ð13Þ

where, Qe, Qt, and t have the same meanings as explained above. The above equation was integrated with boundary conditions, t = 0 with Qt = 0 and t = t with Qt = Qt. The resulting equation is as below. t=Q t ¼ 1=k2 Q e2 þ t=Q t

ð14Þ

In the above equation, k2Q2e was replaced by h, resulting in the following equation.

t=Q t ¼ 1=h þ t=Q e

ð15Þ

0.5 0

ln(Qe-Qt)

0

5

10

15

20

25

30

-0.5 -1 -1.5

-2 -2.5

where, h is initial adsorption rate constant. It was calculated from pseudo-second order plot. Qt/t became h as the time approached zero. k2 is rate constant of pseudo-second order adsorption (g/mg/min). Qe and k2 were calculated from the slope and intercept of the plot t/Qt vs t (Figure not given). The calculated parameters are given in Table 4. The values of k2 and h were 0.016 gmg−1 min−1 and 0.87 gmg−1 min−1. These data established the rapid speed of adsorption initially followed by the slow speed with an increase of time. The regression coefficient (R2 ) was 0.965 high showing the suitability of pseudo-second order model of the adsorption process. But the theoretical and experimental Qe values were not in good agreement in comparison to pseudo-first order kinetic model. Therefore, this model cannot be considered as applicable. 3.10.3. Elovich’s Kinetic Model The rates of adsorption and desorption phenomenon can be established by Elovich’s kinetic model. Formerly Elovich’s equation was developed to illustrate kinetics for gaseous chemo-adsorption on solid surfaces. Afterward, it was modified by Chien and Clayton [29] considering an increase in the activation energy of adsorption linearly with surface coverage. This model is given by the following equation.

calculated by scheming ln(Qe-Qt) vs t plot (Fig. 6) at 293 K temperature. The pseudo-first order rate constant was 0.070 (min−1) with 0.931 as regression coefficient (R2); confirming the fitness of pseudo-first order model. The experimental and theoretical values of Qe were 1.44 and 1.66 mgg−1, respectively (in good agreement). Thus, the experimental data followed well pseudo-first order kinetic model.

dQ t =dt ¼ k2 ðQ e −Q t Þ2

463

35

dQ t =dt ¼ α expð−βQ t Þ

ð16Þ

where, α (mg/g.min) correlates initial adsorption rate. β is desorption constant (g/mg). The above equation was integrated using the boundary conditions, t = 0 with Qt = 0 and t = t with Qt = Qt. Further assuming αβt N N 1, Eq. (16) changes to the following one. Q t ¼ βlnðαβÞ þ βlnt

ð17Þ

The experimental data was applied to Elovich’s model. The values of α, β and R2 are given in Table 4. The values of α, β and R2 were 5.07, mgg−1 min−1, 0.52 gmg−1 and 0.916, respectively. These values were symptomatic of higher adsorption rate than desorption one. Furthermore, the value of regression constant was close to one, confirming the suitability of this model. 3.11. Adsorption mechanism Normally, adsorption progresses by film diffusion, pore diffusion and intra-particle diffusion mechanism. The slowest step is rate determining one, which controls adsorption phenomenon. The experimental data was fitted to intraparticle and film diffusion models to assess the mechanism of fluoride sorption on the reported adsorbent. 3.11.1. Intraparticle Diffusion Kinetic Model The adsorption of adsorbate on adsorbent surface contains different processes such as i) transportation of adsorbate from bulk solution by liquid film to adsorbent surface, ii) sorption of adsorbate on adsorbent surface and iii) move of adsorbate within the pores of adsorbent. Thus, adsorption mechanism is controlled either by the surface adsorption kinetics or transport phenomenon (film and intraparticle diffusions) mechanism or by both processes. The second step is very fast and cannot be rate determining. First and third steps may be the rate determining ones. Thus, these steps were studied by two models. The transfer of fluoride from solution to adsorption sites might be calculated by the relationship between the amount of adsorbed fluoride and the square root of contact time. The equation for this is given below. Q t ¼ kipd t05

ð18Þ

Time (min.)

(a) Fig. 6. The plots are showing pseudo-first-order kinetic plot.

If a graph drawn Qt vs t0.5 is a straight line passing through the origin and slope of the line correlates to rate constant (kipd) the adsorption is controlled by intra-particle diffusion. This graph was drawn (Figure not given). The value of rate constant was calculated to be

464

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

Fig. 7. Mechanism of fluoride adsorption.

0.28 mgg−1 min−0.5. The values of intercept and regression coefficient were 0.266 and 0.991 respectively (Table 4). The graph could not pass via the origin; confirming no applicability intra-particle diffusion model. 3.11.2. Liquid Film Diffusion Kinetic Model Liquid film diffusion kinetics model was developed by Boyd et al. [30]. In this model, boundary plays a noteworthy role in the adsorption process. Liquid film diffusion model is given by the following equation.

ln ð1−Q t =Q e Þ ¼ −k fd t

ð19Þ

5. Desorption Studies To make the method economic desorption and recycling of adsorbent were evaluated. The adsorbent was treated with acids and bases. It was observed that desorption was maximum 20% with acids. However, up to 90% desorption was achieved with strong base sodium hydroxide (1.0 N). The other bases such as potassium hydroxide, calcium hydroxide, and magnesium hydroxides were also used, but the maximum desorption occurred with sodium hydroxide only. The regenerated adsorbent was used for consecutive five cycles for the removal of fluoride. The removal capacities for five cycles ranged from 84-89%. 6. Application of the developed method in real water samples

Or ln ð1−FÞ ¼ −k fd t

ð20Þ

where, F (Qt /Qe) is the fractional attainment of equilibrium. kfd is film diffusion rate constant. If a graph of ln (1-F) vs t is a straight line with zero intercepts the adsorption is supposed to proceed via the film diffusion mechanism. The values of film diffusion rate constant, intercept and regression constant (R2) were 0.071 gmg−1, 0.14 and 0.939, respectively (Table 4). The straight line conceded through the origin with a small digression from zero intercept (− 0.14). This divergence from zero might be due to the high rate of agitation used in the kinetic experiments. Additionally, the disparity between the rate of mass transfer in the preliminary and final steps of adsorption process might be accountable for a small deviation from zero. The analogous results are also available into the literature [31–33]. Thus, fluoride adsorption on the iron nano-impregnated adsorbent was restricted by the liquid film diffusion mechanism. 4. Mechanism of Adsorption at supra-molecular level Generally, nanoparticles are good adsorbents for the removal of various pollutants. Moreover, the reported nano iron-impregnated particles are positively charged. Therefore, these have good attractive tendencies to adhere fluoride ions on their surface. The schematic representation of fluoride removal is shown in Fig. 7. It is clear from this figure that nano iron-impregnated particles have positive charges with bromide as negative counter ions. The electro negativities of fluoride and bromide are 3.98 and 3.16, respectively. Hence, fluoride replaced bromide easily due to its higher electro negativity than bromide. Therefore, the removal of fluoride on the reported adsorbent is controlled by electrostatic forces of attraction. Hence, fluoride replaced bromide easily. That is why the reported method is fast for the removal of fluoride from water.

The developed adsorption method was applied to remove fluoride from the ground water sample. Ten water samples were collected from different places of Ajmer district of Rajasthan, India. Fluoride concentration was determined and ranged from 1.5 to 4.0 mg/L. The developed method was applied to the real life samples for the removal of fluoride. The procedure was as described in the experimental section. It was observed that the percentage removal of fluoride ranged from 90 to 100%. These part of studies clearly indicated that the developed adsorption method is applicable for the removal of fluoride from natural ground water. 7. Conclusion The prepared iron nano-impregnated material (adsorbent) was able to remove fluoride from the water effectively. The adsorption method was selective for fluoride. The percentage removal of fluoride from water was 90%. The adsorption process was exothermic in nature with the extent of fluoride removal in the order of 293 N 298 N 303 K. The adsorption obeyed Langmuir, Freundlich and Temkin models. Thermodynamic studies also proved adsorption exothermic in nature. Kinetic modeling confirmed pseudo-first-order and liquid film diffusion mechanisms. In a nut shell, the presented adsorption method is rapid, eco-friendly and cost-effective as may be applied at pHs of natural water resources; with low contact time and dose regimen. Small contact time and dose regimens are the best attributes of this system making the results manageable at column operation. Thus, the described method may be used for the subtraction of fluoride from any water resource at huge scale economically. Acknowledgement The authors extend their sincere appreciation to the Deanship of Scientific Researchat King Saud University for its funding to this Prolific Research Group (PRG-1436-04).

I. Ali et al. / Journal of Molecular Liquids 211 (2015) 457–465

References [1] V.K. Gupta, I. Ali, V.K. Saini, Water Res. 41 (2007) 3307–3316. [2] WHO, Technical Report Series, 1994. [3] P.T.C. Harrison, Fluoride in water: a UK perspective, J. Fluor. Chem. 126 (2005) 1448–1456. [4] I. Ali, V.K. Gupta, Nature (London) 1 (2006) 2661. [5] I. Ali, Sep. Purif. Rev. 39 (2010) 95–171. [6] I. Ali, Chem. Rev. 112 (2012) 5073–5091. [7] I. Ali, T.A. Khan, M.J. Asim, Environ. Manag. 113 (2012) 170–183. [8] I. Ali, Sep. Purif. Rev. 43 (2014) 175–2015. [9] V.K. Gupta, S.K. Srivastava, D. Mohan, S. Sharma, Waste Manag. 17 (1998) 517–522. [10] V.K. Gupta, S. Agarwal, T.A. Saleh, J. Hazard. Mater. 185 (2011) 17–23. [11] V.K. Gupta, I. Ali, T.A. Saleh, RSC Adv. 2 (2012) 6380–6388. [12] V.K. Gupta, R. Jain, S. Agarwal, M. Shrivastava, Mater. Sci. Eng. C31 (2011) 1062–1067. [13] V.K. Gupta, A. Nayak, Chem. Eng. J. 18 (2012) 081–090. [14] T.A. Saleh, V.K. Gupta, Environ. Sci. Pollut. Res. 19 (2012) 1224–1228. [15] T.A. Saleh, V.K. Gupta, J. Colloid Interface Sci. 371 (2012) 101–106. [16] H. Khani, M.K. Rofouei, P. Arab, V.K. Gupta, Z. Vafaei, J. Hazard. Mater. 183 (2010) 402–409. [17] R. Jain, V.K. Gupta, N. Jadon, K. Radhapyari, Anal. Biochem. 407 (2010) 79–88.

[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

465

T.A. Saleh, V.K. Gupta, J. Colloid Interface Sci. 371 (2012) 101–106. J. Cheng, X. Meng, C. Jing, J. Hao, J. Hazard. Mater. 278 (2014) 343–349. J. Du, D.A. Sabatini, E.C. Butler, Chemosphere 101 (2014) 21–27. Y. Duan, C. Wang, X. Li, W. Xu, J. Water Health 12 (2014) 715–721. W. Ma, T. Lv, X. Song, Z. Cheng, S. Duan, G. Xin, F. Liu, D. Pan, J. Hazard. Mater. 268 (2014) 166–176. A. Teutli-Sequeira, M. Solache-Ríos, V. Martínez-Miranda, I. Linares-Hernández, J. Colloid Interface Sci. 418 (2014) 254–260. Y. Yu, C. Wang, X. Guo, J. Paul Chen, J. Colloid Interface Sci. 441 (2015) 113–120. Z. Wan, W. Chen, C. Liu, Y. Liu, C. Dong, J. Colloid Interface Sci. 443C (2014) 115–124. APHA, Standards Method for the Examination of Water and Waste Water, 1998. G.E. Hoag, J.B. Collins, J.L. Holcomb, J.R. Hoag, M.N. Nadagouda, R.S. Varma, J. Mater. Chem. 19 (2009) 8671–8677. T. Shahwan, S. AbuSirriah, M. Nairat, E. Boyac, A.E. Eroglu, T.B. Scott, K.R. Hallam, Chem. Eng. J. 172 (2011) 258–266. S.H. Chien, W.R. Clayton, Soil Sci. Soc. Am. J. 44 (1980) 265–268. G.E. Boyd, A.W. Adamson, L.S. Myers, J. Am. Chem. Soc. 69 (1947) 2836–2848. C.W. Cheung, J.F. Porter, G. McKay, Water Res. 35 (2001) 605–612. S. Goswami, U.C. Ghosh, Water SA 31 (2005) 57–602. M. Onyango, H. Matsuda, T. Ogada, J. Chem. Eng. Jpn. 36 (2003) 477–485.