Controllable synthesis of mesoporous alumina with large surface area for high and fast fluoride removal

Author’s Accepted Manuscript Controllable synthesis of mesoporous alumina with large surface area for high and fast fluoride removal Naicai Xu, Zhong Liu, Yaping Dong, Tianzeng Hong, Li Dang, Wu Li www.elsevier.com/locate/ceri

PII: DOI: Reference:

S0272-8842(16)31007-0 http://dx.doi.org/10.1016/j.ceramint.2016.06.164 CERI13180

To appear in: Ceramics International Received date: 11 June 2016 Revised date: 23 June 2016 Accepted date: 24 June 2016 Cite this article as: Naicai Xu, Zhong Liu, Yaping Dong, Tianzeng Hong, Li Dang and Wu Li, Controllable synthesis of mesoporous alumina with large surface area for high and fast fluoride removal, Ceramics International, http://dx.doi.org/10.1016/j.ceramint.2016.06.164 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. 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.

Controllable synthesis of mesoporous alumina with large surface area for high and fast fluoride removal Naicai Xu a,b,c , Zhong Liu *a, Yaping Dong a, Tianzeng Hong a,c, Li Dang a,c, Wu Li *a

a. Key Laboratory of Comprehensive and Highly Efficient Utilization of Salt Resources, Qinghai

Institute of Salt Lakes, Chinese Academy of Sciences, Xining, 810008, China

b. Department of Chemistry, Qinghai Normal University, Xining, 810008, China

c. University of Chinese Academy of Sciences, Beijing, 100049, China *

Corresponding author

Qinghai Institute of Salt Lakes, Chinese Academy of Sciences, Xining 810008, China

Email address: [email protected] (Zhong Liu); [email protected] (Wu Li)

Tel: +86-971-636-3311; Fax: +86-971-631-0402.

Abstract

Gamma phase of mesoporous alumina (MA) with large surface area was successfully synthesized

by a facile hydrothermal method followed by thermal treatment for fluoride removal. The as-

synthesized MA nanoparticles with average size of 20 nm−150 nm have ordered wormhole-like

mesoporous structure. The pore size is 5 nm with a narrow distribution, and the specific surface area reaches 357 m2 g-1 while the bulk density is 0.45 cm3 g-1. Glucose as a small-molecule template plays

an important role on the morphology, surface area and pore diameter of the MA. As an ionic adsorbent 1

for fluoride removal, the maximum adsorption capacity of MA is 8.25 mg g-1, and the remove

efficiency reaches 90% in several minutes at pH of 3. The Langmuir equilibrium model is found to be

suitable for describing the fluoride sorption on MA and the adsorption behavior follows the pseudo-

second-order equation well with a correlation coefficient larger than 0.99. The larger surface area and

relatively narrow pore size of MA are believed to be responsible for improving the adsorption

efficiency for fluoride in aqueous solution.

Key words: alumina; glucose; porous material; adsorption; fluoride removal

1. Introduction

Fluoride is one of the essential trace elements for growth and development in human body, which

has important effect on bone metabolism. However, excessive intake of fluoride will lead to many bone

diseases such as brittle bones, osteoporosis, arthritis, cancer and mottling of the teeth, etc. [1-2]. In

addition, high concentration of fluoride in groundwater is a worldwide problem, which is not only

detrimental to human health but also brings serious environmental pollution [3-4]. Therefore, it is

necessary to find out the suitable methods to remove excess fluoride ions in water. The traditional

techniques to remove fluoride include chemical precipitation, membrane processes, ion exchange and

adsorption [5-6]. Among various methods for defluorination of water, adsorption is an ideal and the

most promising method compared to other techniques due to the convenience of operation, lower cost

and being a relatively more environment friendly process [7-8]. As a kind of traditional material, MAs

have considerable applications in catalysis, adsorption, ceramics, heat insulating and sensing fields [5, 2

9-11] owing to their large surface area, unique pore size distribution, acid-base properties and low price.

Using as a significant adsorbent, the adsorption performances of alumina largely depend upon their

crystalline structures and structural properties. Therefore, it is crucially important to prepare alumina

with excellent pore structures. Although researchers have fabricated lots of high efficient metal

composites for defluorination recently, such as Fe-Mg-La [12], Ce-Fe [13], Ca-Al-La [14], Al-Zr [15],

the preparation processes are always complicated and cost is quite high. Therefore, to prepare and

regulate the microstructure of alumina ceramics is still the focus of current research.

At present, a considerable volume of MAs have been successfully synthesized by adopting

nanocasting method [16-17], solvent-deficient synthesis technique [18-19] and sol–gel process [20].

Great efforts have also been devoted to synthesizing MAs by employing surfactants as templates. For

example, Wu et al [21]. synthesized the MA with tunable structural properties via hydrolysis of

aluminium isopropoxide associated with non-ionic block copolymer P123 as the structure-directing

template. Itoh et al [22]. developed the polyol synthesis of γ-alumina particles utilized flux reaction of

ethylene glycol

with aluminum nitrate hexahydrate and poly (vinylpyrrolidone) (PVP). MAs

synthesized by template methods have larger surface area as well as well-organized mesopore structure

[23-24], which can greatly enhance the adsorption capacity of fluoride ions in water. However, the

surfactants introduced into the synthesis process are always expensive organic macromolecular

compounds, which are complicated in the treating process and serious environmental hazardous.

Nowadays, some small molecules templates have attracted more attention to fabricate micro- and 3

mesoporous materials due to their facility of preparation and surface modifications [25]. More than that,

they also can construct multistage orderly pore structures and regulate the structures, morphologies and

surface chemical properties of materials effectively. For instance, saccharide molecules have been

utilized to prepare mesoporous silica [26]. To our best knowledge, there are few reports to study the

effect of glucose on the structural properties of alumina ceramics materials.

In the present work, we carried out the synthesis of MA by hydrolyzing inorganic Al resources in

aqueous solution using glucose as a small molecule template. The resultant sample has large surface

area and ordered wormhole-like pore structures. The sorption experiments like adsorption kinetics,

adsorption isotherm and the effects of pH on fluoride removal have been investigated in a batch model.

2 Experimental

1.1 Chemicals

Glucose monohydrate (C6H12O6·H2O, CAS: 14431-14-7), sodium metaaluminate (NaAlO2, CAS:

1138-49-1) and aluminum chloride hexahydrate (AlCl3·6H2O, CAS: 7784-13-6) were purchased from

Simopharm Chemical Reagent Co., Ltd (China). Sodium fluoride (NaF, CAS: 7681-49-4) was obtained

from Aladdin Industrial Corporation, and absolute ethyl alcohol (C2H5OH, CAS: 64-17-5) was

provided by Tianjin Yongda Chemical Reagent Co., Ltd (China). All chemicals were of analytical

purity (except NaAlO2, CP) and used as received without further purification. Deionized water was

used for all experiments.

1.2 Material Preparation 4

Deionized water of 70 mL was added to 5.0 mmol glucose under stirring in a beaker. After 10 min,

10 mmol AlCl3·6H2O and 10 mmol NaAlO2 were added successively and the mixture was stirred

continuously for another 30 min. Then the mixture was transferred into a 100 mL Teflon lined steel

autoclave and maintained at 150°C for 12 h. After being cooled to ambient temperature, a brown solid

was obtained and washed with deionized water and anhydrous ethanol for several times and finally

dried in an oven at 80°C overnight.

The dried as-synthesized product was further calcined for 4 h at a 2.2°C/min heating rate in

Muffle furnace at 550°C under air atmosphere to get γ-Al2O3 (CAS: 1344-28-1) adsorbent. In order to

study the effects of quantity of glucose on the structural properties of MA, a series of contrast

experiments were also carried out following the above procedures.

1.3 Adsorption tests The fluoride stock solution of 100 mg L-1 was prepared by dissolving 0.2210 g NaF in 1000 mL

deionized water and stored in a polypropylene bottle. Working fluoride anions solutions were obtained

by diluting the stock solution to a certain concentration with deionized water suitably. To test the effect of pH on fluoride removal, 50 mL of 60 mg L-1 sodium fluoride was mixed with

250 mg as-synthesized MA in a 100 mL polyethylene plastic bottle, capped tightly and placed on an

automatic shaker (Aohua THZ-82A, Changzhou, China) under shaking at 25°C. Then the solution was

filtered and the fluoride concentration was analyzed by an ion chromatography (ISC-5000+, Fisher

Scientific, USA). The pH of solution was adjusted by adding 0.1 M NaOH (CAS: 1310-73-2) or 0.1 M 5

HCl (CAS: 7467-01-0) solution, and the initial and final pH values were measured by pH meter

(SevenMulti, Mettler Toledo, Switzerland). For adsorption kinetic tests, 250 mg of as-synthesized MA was added to 50 mL of 60 mg L-1

sodium fluoride (without pH adjustment) under stirring for a specified time. The experimental data

obtained from batch experiments were fitted to pseudo-first-order and pseudo-second-order kinetic

models. Isothermal adsorption experiments were conducted by varying the concentration of sodium fluoride (without pH adjustment) from 10 to 100 mg L-1. The polyethylene plastic bottles were kept in

an automatic shaker for 4 h to reach the equilibrium of the solution with the solid mixture. Langmuir

and Freundlich isotherm models were used to fit the equilibrium data of fluoride adsorption on the

samples.

1.4 Materials Characterization

XRD diffraction patterns of the samples were recorded on X’pert Pro X–ray diffractometer

(Philips, the Netherlands) to identify their crystal structures. Copper Kα radiation (λ = 0.15406 nm) was used with a power setting of 40 kV and 35 mA (Scan rate = 5° min-1). SEM tests were performed on a

JSM-5610LV/INCA microscope (JEOL, Japan) with Au-sputtered specimen operated at 15 keV to

observe the morphologies of the samples. TEM photographs were obtained on a JEM-2100 microscope

(JEOL, Japan) operated at 200 kV. Samples were finely ground in an agate mortar to fine particles and

dispersed ultrasonically in ethanol. The well dispersed samples were deposited on a Cu grid covered by

a holey carbon film for measurements. Nitrogen adsorption-desorption isotherms were conducted on a 6

Quantachrome Autosorb-iQ apparatus (Quantachrome, USA) at -196°C. The specific surface areas

were calculated by using the Brunauer-Emmett- Teller (BET) method over the relative pressure range

of 0.05–0.30, and the pore size distributions (PSD) were calculated from the adsorption branch of the

isotherm with the Barrett-Joyner- Halenda (BJH) model.

3 Results and discussion

3.1 Adsorbent characterization

The crystalline phase of as-synthesized MA is identified by XRD analysis. Figure 1 shows the

XRD patterns of the prepared Al2O3. The diffraction peaks at 2θ of 67°, 60.9°, 45.8°, 39.5°, 37.6°, 31.9°

and 19.4° correspond to the reflection from (440) , (511), (400) , (222) , (311) , (220) and (111) planes

in cubic phase of γ-Al2O3 (JCPDS-10-0425) [27], respectively. No other diffraction peaks representing

other phases were detected, which indicates a high purity of the prepared samples. In addition, when

glucose was not added to the reaction system, the γ-Al2O3 displays a relatively sharper diffraction

peaks (shown in Fig. 1a), which presents a better crystallinity than other aluminas templated by glucose.

The results show that the amount of glucose almost has no effect on the crystalline phase and the

crystalline degree of gamma-alumina.

Fig. 2 shows the SEM images of γ-Al2O3 prepared by adding a certain amount of glucose. When

glucose was not added to the reaction system, gamma alumina presents irregular and agglomerated

nanoparticles morphology and mingles with some tiny fibers (Fig. 2a). While a series of glucose with

different concentration (1.0 mmol, 2.5 mmol, 4.0 mmol and 5.0 mmol) was added to the double 7

hydrolysis reaction system, the final products take on various morphologies as following: sheet-like

(Fig. 2b), flower-like (Fig. 2c), soil bulks (Fig. 2d) and colloidal state with a rough surface (Fig. 2e)

successively. In addition, the TEM images of γ-Al2O3, embedded in Fig. 2c and Fig. 2e respectively

show a typical wormhole-like morphology known for mesoporous structure without any evident

features [28]. The results demonstrate that glucose has significant effect on the morphologies of γ-

Al2O3, and these small molecule templates rebuilt the structural properties of MA to a great degree. Fig. 3 presents nitrogen adsorption–desorption isotherms and corresponding PSD curves of γ-

Al2O3. We can see that all the samples exhibit typical IV isotherm with H1 hysteresis loops (Fig. 3A)

according to the IUPAC classification [29], indicating their uniform mesopore structures. The well-

formed H1 hysteresis loop is regarded as associating with the capillary condensation in large pore

channels with probable channel adjustment [21, 30]. At same time, we also detect a very obvious shift

of the capillary condensation steps towards smaller relative pressures except MA-0, indicating a

substantial reduce in the pore size of MA. As can be seen from Fig.3 B, MA-0 has a larger pore

diameter (9.7 nm) than other aluminas (4.5 and 5 nm). The changes of pore diameter are in accordance

with the above analysis results, suggesting our deduction and conclusion are reasonable. Structural

properties of the corresponding MAs are presented in Table 1. The specific surface area of MA-0 is 191 m2 g-1, which is smaller than other MAs templated by glucose (more than 300 m2 g-1). In addition,

the pore volumes of samples are also affected by glucose added to the reaction according to Table 1.

By comparison, we can see that the specific surface area of MA significantly increased with addition of 8

glucose to the reaction system, which confirm that small molecule template glucose can contribute to

constructing orderly porous structures and regulating structural properties of MA effectively.

3.2 Possible formation process for mesoporous structure of MA

As shown in Fig. 4, in order to explain the effect of glucose on the pore structures of mesoporous

alumina, we present a schematic diagram to illustrate the possible formation process. First, inorganic

Al resources dissolved in aqueous solution, then formed nucleus gradually and chemically reacted with

each other under certain temperature and pressure. Second, small molecules glucose started to interact

with produced alumina and unreacted Al sources to generate a coating material. The shapes of the

ultimate sample mostly like an egg, the inner substance is glucose and the outer material is something

about aluminum. Finally, the coating material began to decompose under the high temperature and

pressure. Both glucose and water molecules were removed to produce alumina with orderly small pores

under continue heating. This critical process leads to creating a high specific surface area of gamma-

alumina. However, if an organic macromolecule was chosen to act as a template, the pore size of MA

produced eventually is obvious larger than those templated by glucose, but the surface area is relatively small (< 200 m2 g-1). From what have been mentioned above, we can conclude that the structure

directing agent has a significant effect on the structural properties of MA.

3.3 Adsorption kinetic studies

Fig. 5(a) shows the adsorption kinetic curve of fluoride removal over MA-glucose (5.0 mmol) adsorbent at fluoride initial concentration at 60 mg L-1 (without pH adjustment). It can be observed that 9

most of fluoride ions were removed within 5 min, the rate of adsorption had some ups and downs to

some degree after the initial 5 min, and the residual fluoride concentration in solution reached an

almost constant value until the adsorption time reached 180 min. Tang et al. [31] once reported that the

adsorption equilibrium time of removing fluoride by alumina adsorbent is approximately 600 min. In

our work, fluoride ions adsorbed on MA occurred rapidly and reached equilibrium within 180 min. The

higher adsorption rate mainly can be attributed to larger specific surface area and greater accessibility

to pores of alumina adsorbent [32-34].

The pseudo-first-order and pseudo-second-order adsorption models [34-35] were employed to fit

the adsorption kinetic data. The pseudo-first-order adsorption kinetic equation is given as:

ln(qe  qt )  ln qe  k1t

(1)

The pseudo-second-order adsorption kinetic equation is given as:

t 1 1   t 2 qt k2 qe qe

(2)

Where k1 is the pseudo-first-order rate constant of adsorption (min-1); k2 is the pseudo-secondorder rate constant of adsorption (min g-1 mg-1); qe (mg g-1) and qt (mg g-1) are the amounts of fluoride

adsorbed reaching to equilibrium adsorption capacity and at a certain time t (min), respectively.

Kinetic parameters of pesudo–first–order and pseudo–second–order for the adsorption fluoride

ions onto MA were shown in Table 2. Meanwhile, we also represent the pseudo–second–order linear

plot for the MA as shown in Fig. 5b, where the kinetic data were plotted for t/qt against t. Table 2 and

Fig. 5b indicate that the kinetic data fit well with the pseudo–second–order kinetic model. The value of 10

correlation coefficient R2 for pseudo–second–order kinetic model is relatively high (more than 0.99),

and the adsorption capacity of MA for fluoride calculated by this model is also close to that obtained from experiments (10.3701 mg g-1). Compared with pseudo–second–order kinetic model, the predicted equilibrium adsorption capacity (9.8677 mg g-1) of the pseudo–first–order kinetic model is slightly lower than the experimental values (10.3701 mg g-1), and the value of correlation coefficient R2 is

0.9766. From above results we can conclude that both the pseudo–first–order and pseudo–second–

order kinetic models can well describe the adsorption kinetic data in our work. The well-fitting to

pseudo–second–order kinetic model suggests that the fluoride adsorption on MA follows

chemisorptions process involving ion exchange [33].

3.4 Adsorption isotherm studies

Adsorption capacity is one of the most important parameter for adsorbent itself and the design of

all the adsorption process. In order to obtain the maximum adsorption capacity and the optimum

conditions for fluoride ions removal, two most widely used equilibrium models were employed to carry

out the equilibrium fluoride adsorption data (qe VS. Ce). One is Langmuir, which indicates the

monolayer adsorption on uniform homogeneous surface with identical sites of adsorbent [36]; the other

one is Freundlich, which indicates the heterogeneity of the adsorbent surface and considers multilayer

adsorption [37].

The liner equation of Langmuir model is presented as:

Ce 1 1   Ce qe Qmax b Qmax

(3)

11

Where qe is the equilibrium adsorption capacity (mg g-1); Ce is the equilibrium ions concentration in aqueous solution (mg L-1); Qmax is the maximum adsorption capacity (mg g-1); b is the Langmuir adsorption constant (L mg-1).

The liner equation of Freundlich model is presented as:

1 ln qe  ln K F  ln Ce n

(4)

Where qe is the equilibrium adsorption capacity (mg g-1); Ce is the equilibrium ions concentration in aqueous solution (mg L-1), KF is the Freundlich constant (L mg-1), and n is the adsorption intensity

(heterogeneity) factor.

The results of equilibrium adsorption capacity of fluoride (qe VS Ce) were shown in Fig. 6. It can be seen clearly that when the adsorption data for fluoride original concentration range (10–100 mg L−1)

were considered, the equilibrium adsorption capacity increased with increasing the equilibrium concentration of fluoride from 0.69 mg L-1 to 40 mg L-1. When the fluoride equilibrium concentration exceeds 40 mg L-1, the adsorption capacity of fluoride no longer increased and the corresponding curve

became smooth relatively. Fig. 7 shows the curves of adsorption isotherm of Langmuir and Freundlich

models. It can be observed that all the equilibrium adsorption capacity increased with increasing the

concentrations of fluoride until the adsorption process reached an equilibrium state. The fitted

parameters obtained from the two models are given in Table 3. It is obvious that Langmuir model gives

a better fit to the experimental data with a correlation coefficient of 0.984, while the correlation

coefficient of Freundlich model is only 0.785. The results indicate that Freundlich model is unsuitable 12

to evaluate the fluoride adsorption process, and the process of adsorption fluoride on MA involves the

monolayer coverage of fluoride on the surface of the adsorbent at specific sites.

It is noted that Weber and Chakravotri [38] once defined the Equilibrium parameter “RL” to

indicate the shape of the isotherm according to the Langmuir model, and the mathematical expression

is shown below:

RL 

1 1  bC0

(5)

Where C0 refers to the initial fluoride concentration and b is Langmuir constant. When the value

of RL is within 0 and 1, it represents the favorable adsorption process [39]. However, if the value of RL

exceeds 1, indicating the adsorption process is unfavorable [40]. RL values calculated from the present

system were presented in Table 3. It can be seen clearly that all the values of RL vary from 0.2301 to 0.0281 with the variation of C0(F-) from 10 to 100 mg L−1. This result indicates that adsorption fluoride

ions onto MA are a favorable process in our work. The kind of situation always emerges at that time

when the fluoride concentration is relatively higher.

3.5 The effect of solution pH on fluoride removal

The solution pH is extremely important for fluoride removal in solid/aqueous separation process,

since it determines the adsorption potential at adsorbent–water interface and is related to the pHzpc (the

pH point of zero charge) of the adsorbents [41]. Therefore, the fluoride adsorption onto MA at pH

values ranging from 3 to 11 was investigated systematically when the initial fluoride concentration was maintained at 60 mg L-1. Fig. 8 presents the effect of pH on fluoride removal. For each initial pH, the 13

final pH is also measured after adsorption equilibrium. The results indicate that the optimum pH for

fluoride adsorption is 3, and the fluoride removal rate can reach 97.7% under the present conditions.

The fluoride removal rate decreases sharply (60%) with increasing solution pH until reaches pH of 4.0.

Thereafter, the fluoride removal efficiency declines persistently until it attains a fixed value

(approximately 50%) at pH of 5. The final equilibrium pH of solution keeps at 7.5–7.7 after adsorption,

which is more neutral than the optimum pH for MA. Finally, the fluoride removal rate falls to 34.39%

when the solution of pH increases to 11. From above results we can see that the as-synthesized MA is

unsuitable to remove fluoride ions in strong alkali solution. The fluoride remove efficiency was closely

depended on the solution pH and the maximum adsorption capacity for fluoride was obtained at pH of

3 in present work. This phenomenon can be explained as follows. On one hand, the fluoride ions can

form a little acid HF or complexation AlFx when the pH of solution is lower than 3, which bring about

less removal efficiency for fluoride. On the other hand, the electrostatic repulsion of fluoride ions to the

negatively charged surface of MA is remarkably increasing when the pH of solution exceeds 3. The

competition for active sites among the excessive hydroxide ions is also notable, which cause a

progressively decrease for fluoride removal with increasing pH of solution. In addition, we can see

from secondary axis that every equilibrium pH values of the solution after sorption fluoride is

wandering around 7 except both ends values of curve (5.9, 8.2 respectively), which is quite close to the

pH of a neutral water. This result indicates that it is unnecessary to adjust pH of solution with

acid/alkali in the process of treating drinking water when the final pH is around 7. This dominant 14

viewpoint will lead to cost reduction and simplify the complicated procedures for removing fluoride in

water.

4 Conclusions

In this study, MAs with different morphologies were successfully synthesized by introducing

small molecule template agent–glucose. The quantity of glucose has a significant effect on the

structural properties of MA. This promising adsorbent material displays a higher and efficient fluoride

removal in water. The main conclusions drawn from the present study are summarized as follows:

(1) Glucose is highly important to prepare of MA with tunable structural properties. However, it

has no effect on both crystal structure and crystalline degree of MA.

(2) The defluoridation speed is very quick during the initial 5 min and slows down gradually until

it reaches the adsorption equilibrium within 180 min. Adsorption kinetics fit pseudo–second–order

model well, suggesting the removal of fluoride ions complies with chemisorptions process involving

ion exchange.

(3) Langmuir adsorption isotherms can predict the adsorption equilibrium data well while

Freundlich mode is unsuitable to evaluate the adsorption process, indicating the adsorption of fluoride

on MA referring to the monolayer coverage at specific sites. The equilibrium parameter RL calculated

in this work implies the adsorption fluoride on MA is a favorable process.

(4) The solution pH has significantly effect on the remove efficiency of fluoride and the maximum

adsorption capacity acquired at pH of 3 for MA. 15

Acknowledgements

This work was financially supported by the National Nature Science Foundation of China (No:

51302280, 21557001) and the Natural Science Foundation in Qinghai province (No: 2014-ZJ-936Q,

2014-ZJ-778).

References:

[1] N. J. Chinoy, Effects of fluoride on physiology of animals and human beings, Indian J. Environ.

Toxicol. 1 (1991) 17-32.

[2] P. T. C. Harrison, Fluoride in water: a UK perspective, J. Fluorine Chem. 126 (2005) 1448-1456.

[3] M. Srimurali, A. Pragathi, J. Karthikeyan, A study on removal of fluorides from drinking water by

adsorption onto low-cost materials, Environ. Pollut. 99 (1998) 285-289.

[4] M. Amini, K. Mueller, K. C. Abbaspour, T. Rosenberg, M. Afyuni, K. N. Møller, M. Sarr, C. A.

Johnson, Statistical modeling of global geogenic fluoride contamination in groundwaters, Environ. Sci.

Technol. 42 (2008) 3662-3668.

[5] A. Bhatngar, E. Kumar, M. Sillanpaa, Fluoride removal from water by adsorption—a review, Chem.

Eng. J. 171 (2011) 811–840.

[6] N. Azbar, A. Turkman, Defluoridation in drinking waters, Water Sci. Technol. 42 (2000) 403–407.

[7] K. Huang, J. G. Shao, H. M. Zhu, K. Inoue, Removal of fluoride from aqueous solution onto Zr-

loaded garlic peel (Zr–GP) particles, J. Cent. South Univ. T. 18 (2011) 1448–1453.

[8] S. P. Kamble, P. Dixit, S. S. Rayalu, N. K. Labhsetwar, Defluoridation of drinking water using 16

chemically modified bentonite clay, Desalination, 249 (2009) 687–693.

[9] S. S. Akarmazyan, P. Panagiotopoulou, A. Kambolis, C. Papadopoulou, D. I. Kondarides, Methanol

dehydration to dimethylether over Al2O3 catalysts, Appl. Catal. B: Environ. 145 (2014) 136–148.

[10] I. Kheirollahi, M. Abdellahi, M. Emamalizadeh, H. Sharifi, Preparation and characterization of

multilayer mesoporous alumina nano membrane via sol–gel method using new precursors, Cream. Int.

41 (2015)15083–15088.

[11] P. Kathirvel, J. Chandrasekaran, D. Manoharan, S. Kumar, Preparation and characterization of

alpha alumina nanoparticles by in-flight oxidation of flame synthesis, J. Alloys Compd. 590 (2014)

341-345.

[12] Y. Yu, L. Yu, P. Chen, Adsorption of fluoride by Fe-Mg-La triple-metal composite: adsorbent

preparation, illustration of performance and study of mechanisms, Chem. Eng. J. 262 (2015) 839-846.

[13] D. D. Tang, G. K. Zhang, Efficient removal of fluoride by hierarchical Ce-Fe bimetal oxides

adsorbent: Thermodynamics, kinetics and mechanism, Chem. Eng. J. 283 (2016) 721-729.

[14] W. Xiang, G. K. Zhang, Y. L. Zhang, D. D. Tang, J. T. Wang, Synthesis and characterization of

cotton-like Ca-Al-La composite as an adsorbent for fluoride removal, Chem. Eng. J. 250 (2014) 423-

430.

[15] J. Y. Zhu, X. Y. Lin, P. W. Wu, Q. S. Zhou, X. G. Luo, Fluoride removal from aqueous solution

by Al(Ⅲ)-Zr(Ⅳ) binary oxide adsorbent, Appl. Surf. Sci. 357 (2015) 91-100.

[16] C. F. Chen, Z. D. Ding, Z. D. Tan, H. H. Qi, Y. H. He, Preparation of nano α-alumina powder and 17

wear resistance of nanoparticles reinforced composite coating, Powder Technol. 257 (2014) 83-87.

[17] Z. X. Wu, Q. Li, D. Feng, P. A. Webley, D. Y. Zhao, Ordered mesoporous crystalline gamma-

Al2O3 with variable architecture and porosity from a single hard template, J. Am. Chem. Soc. 132

(2010) 12042–12050.

[18] B. Huang, C. H. Bartholomew, B. F. Woodfield, Facile structure-controlled synthesis of

mesoporous γ-alumina: Effects of alcohols in precursor formation and calcinations, Microporous

Mesoporous Mater. 177 (2013) 37–46.

[19] B. Huang, C. H. Bartholomew, S. J. Smith, B. F. Woodfield, Facile solvent-deficient synthesis of

mesoporous gamma-alumina with controlled pore structures, Microporous Mesoporous Mater. 165

(2013) 70–78.

[20] K. R. Murali, P. Thiumoorthy, Characteristics of sol–gel deposited alumina films, J. Alloys

Compd. 500 (2010) 93-95.

[21] W. Wu, Z. J. Wan, W. Chen, M. M. Zhu, D. K. Zhang, Synthesis of mesoporous alumina with

tunable structural properties, Microporous Mesoporous Mater. 217 (2015) 12–20.

[22] T. Itoh, T. Uchida, I. MatsuBbara, N. Izu, W. Shin, H. Miyazaki, H. Tanjo, K. Kanda, Preparation

of γ-alumina large grain particles with large specific surface area via polyol synthesis, Cream. Int. 41

(2015) 3631–3638.

[23] A. L. Vazquer, R. Carreraa, E. Arce, M. M. Pineda, Preparation of mesoporous alumina films by

anodization: Effect of pretreatments on the aluminum surface and MTBE catalytic oxidation, J. Alloys 18

Compd. 483 (2009) 418-421.

[24] C. Yang, L. L. Gao, Y. X. Wang, S. Komarneni, Fluoride removal by ordered and disordered

mesoporous aluminas, Microporous Mesoporous Mater. 197 (2014) 156–163.

[25] A. S. Patel, T. Mohanty, Silver nanoclusters in BSA template: a selective sensor for hydrogen

peroxide, J. Mater. Sci. 49 (2014) 2136–2143.

[26] E. Vega, C. Marzabadi, Y. Kazakevich, A. Y. Fadeev, Synthesis of chiral mesoporous silicas with

oligo (saccharide) surfaces and their use in separation of stereoisomers, J. Colloid Interface Sci. 359

(2011) 542-544.

[27] J. R. Ge, K. J. Deng, W. Q. Cai, J. B. Zhou, Effect of structure-directing agents on facile

hydrothermal preparation of hierarchical γ-Al2O3 and their adsorption performance toward Cr (VI) and

CO2, J. Colloid Interface Sci. 401 (2013) 34–39.

[28] G. Lee, C. Chen, S. T. Yang, W. S. Ahn, Enhanced adsorptive removal of fluoride using

mesoporous alumina, Microporous Mesoporous Mater. 127 (2010) 152–156.

[29] G. J. Ji, M. M. Li, G. H. Li, G. M. Gao, H. F. Zou, S. C. Gan, X. C. Xu, Hydrothermal synthesis of

hierarchical micron flower-like γ-AlOOH and γ-Al2O3 superstructures from oil shale ash,

Powder Technol. 215-216 (2012) 54-58.

[30] T. Horikawa, D. D. Do, D. Nicholson, Capillary condensation of adsorbates in porous materials,

Adv. Colloid Interface Sci. 169 (2011) 40-58.

[31] Y. L. Tang, X. H. Guan, T. Z. Su, J. M. Wang, Fluoride adsorption onto activated alumina: 19

Modeling the effects of pH and some competing ions, Colloid Surf. A-Physicochem. Eng. Asp. 337

(2009) 33-38.

[32] R. Krishna, A. V. V. S. Hema Swamy, Physico-chemical key parameters, langmuir and freundlich.

Eur. Chem. Bull. 1 (2012) 258-262.

[33] J. Hao, M. Han, C. Wang, X. G. Meng, Enhanced removal of arsenite from water by a mesoporous

hybrid material-Thiol-functionalized silica coated activated alumina, Microporous Mesoporous Mater.

124 (2009) 1–7.

[34] Y. Miyake, H. Ishida, S. Tanaka, P. S. Kolev, Theoretical analysis of the pseudo-second order

kinetic model of adsorption. Application to the adsorption of Ag (I) to mesoporous silica microspheres

functionalized with thiol groups, Chem. Eng. J. 218 (2013) 350-357.

[35] Y. S. Ho, Review of second-order models for adsorption systems, J. Hazard. Mater. 136 (2006)

681-689.

[36] D. Marani, B. R. Sudireddy, L. Nielsen, S. Ndoni, R. Kiebach, Poly(vinylpyrrolidone) as

dispersing agent for cerium-gadolinium oxide (CGO) suspensions, J. Mater. Sci. 51 (2016) 1098-1106.

[37] N. Salahudeen, A. S. Ahmed, A. H. Al-Muhtaseb, M. Dauda, S. M. Waziri, B. Y. Jibril, J. Al-

Sabahi, Synthesis, characterization and adsorption study of nano-sized activated alumina synthesized

from kaolin using novel method, Powder Technol. 280 (2015) 266–272.

[38] T. W. Weber, R. K. Chakravorti, Pore and solid diffusion models for fixed-bed adsorbers, AIChE.

J. 20 (1974) 228–238. 20

[39] W. Nigussie, F. Zewge, B. S. Chandravanshi, Removal of excess fluoride from water using waste

residue from alum manufacturing process, J. Hazard. Mater. 147 (2007) 954–963.

[40] S. Jagtap, M. K. N. Yenkie, N. Labhsetwar, S. Rayalu, Defluoridation of drinking water using

chitosan based mesoporous alumina, Microporous Mesoporous Mater. 142 (2011) 454–463.

[41] H. Y. Jin, Z. J. Ji, J. Yuan, J. Li, M. Liu, C. H. Xu, J. Dong, P. Hou, S. Hou, Research on removal

of fluoride in aqueous solution by alumina-modified expanded graphite composite, J. Alloys Compd.

620 (2015) 361-367.

21

Tables and figure captions

Table 3 Langmuir and Freundlich isotherm constants for adsorption of fluoride onto MA adsorbent tested (initial fluoride concentrations is 10–100 mg L-1, without pH adjustment). Fig. 1 XRD patterns of MAs prepared by adding different quantity of glucose: (a) MA–0; (b) MA–1.0; (c) MA–2.5; (d) MA–4.0 and (e) MA–5.0 (unit: mmol). Fig. 2 SEM images of MAs prepared by adding different quantity of glucose: (a) MA–0; (b) MA–1.0; (c) MA–2.5; (d) MA–4.0 and (e) MA–5.0 (unit: mmol). Inset shows corresponding TEM images. Fig. 3 N2 adsorption-desorption isotherms (A) and corresponding pore size distributions (B) of MAs prepared by introducing of different quantity of glucose: (a) MA–0; (b) MA–1.0; (c) MA–2.5; (d) MA– 4.0 and (e) MA–5.0 (unit: mmol). Fig. 4 Scheme representation of the pore size of mesoporous alumina tuned by adding glucose and organic macromolecules. Fig. 5 (a) Adsorption kinetic curves of fluoride ions removal and (b) pseudo–second–order adsorption rates of fluoride over MA sample tested (dosage = 5 g L-1, C0 = 60 mg L-1, r = 150 rpm, T = 25 °C). Fig. 6 Fluoride adsorption isotherm on MA adsorbent tested (dosage = 5 g L-1, C0 = 10–100 mg L-1, r = 150 rpm, T = 25 °C). Fig. 7 The Langmuir (A) and Freundlich (B) isotherm model plots for F - ions adsorption onto the MA adsorbent (dosage = 5 g L-1, C0 = 10–100 mg L-1, r = 150 rpm, T = 25 °C). Fig. 8 Effect of the initial pH on the percentage removal of fluoride by MA adsorbent (dosage = 5 g L -1, C0 = 60 mg L-1, r = 150 rpm, T = 25 °C); secondary axis: variation of equilibrium solution pH versus initial pH.

22

Table 1 Structural properties of as-synthesized MA samples.

BET surface

Adsorbent

Pore volume

2

Average pore dimeter (nm)

area (m /g)

(cm3/g)

MA–0

191

0.54

9.7

MA–glucose(1.0 mmol)

437

0.60

5.0

MA–glucose(2.5 mmol)

321

0.43

5.0

MA–glucose(4.0 mmol)

359

0.44

4.5

MA–glucose(5.0 mmol)

357

0.45

5.0

Experimental

Pseudo-first-order

Pseudo-second-order

qe(mg g-1)

k1(min-1)

qe(mg g-1)

R2

k2(g mg-1 min-1)

qe(mg g-1)

R2

10.3701

0.9388

9.8677

0.9766

0.022

10.3124

0.9976

Table 2 Kinetic paraments of pesudo-first-order and pseudo-second-order for the adsorption of fluoride onto MA adsorbent tested.

Table 3 Langmuir and Freundlich isotherm constants for adsorption of fluoride onto MA adsorbent tested (initial fluoride concentrations is 10–100 mg L-1, without pH adjustment). T(˚C)

Langmuir parameters -1

-1

n

KF(mg g-1)(L mg-1)1/n

R2

0.984

4.055

3.068

0.785

20

40

60

80

100

0.1262

0.0673

0.0459

0.0348

0.0281

Qmax(mg g )

KL(L mg )

8.25

0.346

C0 (mg L-1)

10

RL

0.2301

25

Freundlich parameters 2

R

23

24

25

26

27

28

29

30

31