论文部分内容阅读
Abstract： Fe-Al （hydr）oxide nano-/micro-particles were well grown and dispersed on a wheat straw template， which was characterized by a scanning electron microscope （SEM） with energy-dispersive X-ray spectroscopy （EDS）， X-ray diffraction （XRD）， and a vibrating sample magnetometer （VSM）. The adsorption mechanism of the biomass-based Fe-Al （hydr）oxide nanocomposite was studied by the adsorption isotherms， which followed the Langmuir model better than the Freundlich and Temkin models. In particular， a synergistic adsorption by the mixed Fe-Al （hydr）oxide nano-/micro-particles based on the wheat straw was found， with higher maximum adsorption capacity （Q0） than that of the material containing only Fe3O4 or Al（OH）3 nano-/micro-particles， which was most obvious when the mole ratio of Fe to Al was 1：1. The degree of this unusual effect was reasonably determined by the departure between the experimental and calculated maximum adsorption capacity （Q0-Q0（cal））， which showed that the synergistic effect was most pronounced when the mole ratio of Fe to Al was approximately 1：1. The good adsorption capacity of the mixed Fe-Al （hydr）oxide nano-/micro-particles and the good dispersity by the wheat straw matrix were combined in the biomass-based Fe-Al （hydr）oxide nanocomposite. The nanocomposite material showed high adsorption capacity for both fluoride （F-） and arsenic （As（III） and As（V））， and had the advantage of magnetic separation by tuning its compositions.
Keywords： adsorption; synergistic effect; nanocomposite; biomass; fluoride; arsenic
1 Introduction
Anion pollution in water is a worldwide problem harming human health and natural ecosystems[1-6]. Fluoride is a necessary element for humans; however， the excessive ingestion of fluoride may cause many chronic diseases[7-8]. Arsenic， which mainly exists as arsenite and arsenate in nature， can seriously damage human systems and increase cancer risk[9]. In many areas， water resources continuously deteriorate and are possibly contaminated by both fluoride and arsenic[10]. Therefore， the removal of these harmful anions from water is critical.
Many techniques have been developed for the purification of water， such as precipitation[11]， filtration[12-13]， membrane separation[14-15]， coagulation[16-17]， ion exchange[18-19]， oxidation[20-21]， electrolysis[22-23]， and adsorption[24-26]. Among them， adsorption is one of the most important methods because of its wide applicability， easy operation， high efficiency， and low cost. Nanoparticles have attracted significant interest for the adsorption and removal of pollutants from water owing to their high efficiency based on large surface areas. Many nanoparticle adsorbents composed of alumina[27]， iron oxide[28]， magnesium oxide[29]， titanium oxide[30]， and zinc oxide[31] have been applied to water treatment. A general problem of nanoparticle adsorbents is their aggregation， which leads to the decrease in the number of active sites[6]. In addition， the residue of nanoparticles in water could cause a secondary pollution. To avoid the aggregation of nanoparticles， template materials to disperse them， including graphene oxide[32]， microspheres[33-36]， carbon[37-39]， layered double hydroxides[40]， and cellulose[41-42]， have been extensively studied. Wheat straw is an eco-friendly and low-cost biomass-based material available in large amounts， which is mainly composed of cellulose， hemicelluloses， and lignin with numerous hydroxyl groups. Metal ions could be adsorbed by the hydroxyl groups on the surface of wheat straw and grown to nano-/micro-particles[43]. With a vascular bundle structure and hydroxyl groups on its surface[44-45]， wheat straw could be a suitable matrix to disperse nanoparticles. A previous study showed that the aggregation of magnetic nano-/micro-particles can be efficiently avoided by growing and dispersing them onto wheat straw， and the material can be used for adsorption and easily separated from water[43].
In this study， Fe-Al （hydr）oxide nano-/micro-particles were simultaneously grown onto wheat straw through a co-precipitation method. A synergistic adsorption by the mixed Fe-Al （hydr）oxide nano-/micro-particles dispersed onto wheat straw was unexpectedly found with the enhancement of adsorption capacity. The synergistic effect was determined by the departure between the experimental and calculated maximum adsorption capacity fitted from Langmuir adsorption isotherms. The synergistic effect for the adsorption was most pronounced when the Fe and Al mole ratio was approximately 1：1. This material could be used for the removal of fluoride and arsenic and be separated from water under a magnetic field.
2 Experimental
2.1 Materials
Wheat straw was obtained from the countryside of China. After being cleaned and dried， the wheat straw was mechanically ground into fragments by a miniature plant grinding machine （FZ102， Tianjin City Taisite Instrument Co.， Ltd）. The specific surface area of the wheat straw fragments is 1.91 m2/g. Ferric chloride （FeCl3·6H2O）， ferrous sulfate （FeSO4·7H2O）， aluminum nitrate （Al（NO3）3·9H2O）， and ammonia water （NH3·H2O， 25 wt%） were used for the synthesis of Fe-Al （hydr）oxide nano-/micro-particles. Sodium fluoride （NaF）， sodium arsenite （NaAsO2）， and disodium hydrogen arsenate （Na2HAsO4·7H2O） were used for the test solutions. All the chemicals were purchased from Beijing Chemical Works （Beijing， China） with an analytical grade and used without further purification.
2.2 Synthesis of biomass-based Fe-Al （hydr）oxide nanocomposite The biomass-based Fe-Al （hydr）oxide nanocomposite was synthesized through a co-precipitation method. Approximately 0.8 g of wheat straw fragments were suspended in a 50-mL solution with FeCl3， FeSO4 （the mole ratio of Fe3+ to Fe2+ is 2：1）， and Al（NO3）3. The total ion concentration of the solution was 0.5 mol/L， with mole ratios of total Fe to Al set as 4：1， 3：2， 1：1， 2：3， and 1：4. Under nitrogen atmosphere， 10 mL of ammonia water （25 wt%） was added dropwise into the solution with vigorous stirring at 70℃ for 4 h. The synthesized materials were separated and washed by deionized water several times， and then dried for use. Wheat straw-based nanocomposite loaded only with Fe3O4 or Al（OH）3 was also synthesized for comparison （the products named as Fe3O4 nanocomposite and Al（OH）3 nanocomposite， respectively）.
2.3 Characterization of biomass-based Fe-Al （hydr）-oxide nanocomposite
The specific surface area of the wheat straw fragments was measured through nitrogen adsorption by the BET method using an automated gas sorption analyzer （Quadrasorb SI-MP）. The morphologies of the synthesized biomass-based Fe-Al （hydr）oxide nanocomposites were observed by a scanning electron microscope （SEM） （S-4800， Hitachi） equipped with energy-dispersive X-ray spectroscopy （EDS）. X-ray diffraction （XRD） was studied by using a D8 Focus （Bruker） for crystal characterization. The magnetic properties were characterized using a vibrating sample magnetometer （SQUID-VSM， Quantum Design）.
2.4 Adsorption experiments
Adsorption experiments were conducted in an aqueous solution with different initial concentrations of ions under stirring at room temperature for 24 h. The adsorbent dose was 1 g/L for all the experiments. The concentrations of fluoride ion were measured by a fluoride analyzer （PXSJ-216， INESA）. Arsenic was quantified by an inductive coupled plasma emission spectrometer （ICP） （710-OES， Varian）. The adsorption capacity qe （mg/g） was calculated from the following equation （Eq.（1））：
Where C0 （mg/L） and Ce （mg/L） represent the initial and equilibrium concentration of adsorbate in the solution， respectively. V （L） is the solution volume， and m （g） is the adsorbent mass.
3 Results and discussion
3.1 Characterization of the nanocomposite material
The surface morphologies of wheat straw fragments and the nano-/micro-particle-loaded nanocomposites were studied by SEM， as shown in Fig.1. The surface of the original wheat straw fragments is very smooth （Fig.1（a））. Growth of Fe-Al （hydr）oxides （n（Fe）：n（Al）=1：1） results in the flocculent deposition of well dispersed nano-/micro-particles as shown in Fig.1（b）， indicating the suitability of wheat straw as a template. Fig.1（c） and Fig.1（d） show that Fe3O4 nanoparticles are well dispersed by the wheat straw， while the Al（OH）3 particles have almost grown to micrometer size and are coarsely distributed. As shown in Fig.2（a）～Fig.2（c）， the mixed Fe-Al （hydr）oxide nano-/micro-particles （n（Fe）：n（Al）=1：1） could well grow to coral-like floccules， and both Fe3O4 and Al（OH）3 compositions are uniformly distributed on the surface of the wheat straw matrix investigated by EDS elemental mapping. The co-precipitation formation process of Fe-Al （hydr）oxides onto wheat straw is likely to be effective in avoiding the aggregation of particles， as shown in Fig.2（d）.
The XRD patterns of wheat straw fragmets， Fe-Al （hydr）oxide nanocomposite （n（Fe）：n（Al）=1：1）， Fe3O4 nanocomposite， and Al（OH）3 nanocomposite are shown in Fig.3. The diffraction peaks at 2q=16.2° and 22.2° are those of native cellulose I （Fig.3（a））[44]. Loading with Fe-Al （hydr）oxide nano-/micro-particles causes the appearance of many other diffraction peaks （Fig.3（b））， nearly hiding the cellulose peaks. That is because the full growth of Fe-Al （hydr）oxides coats the surface of the wheat straw. In Fig.3（c） and Fig.3（d）， the characteristic diffraction peaks of Fe3O4 and Al（OH）3 are found. The Fe3O4 diffraction peaks are at 2q=30.4°， 35.6°， 43.3°， 57.1°， 62.8°[46] and the typical Al（OH）3 characteristic diffraction peaks are at 2q=13.8°， 28.5°， 38.4°， 49.1°[47]. The diffraction peak at 2q=22.2° of cellulose is relatively more obvious in Al（OH）3 nanocomposite， which is possibly due to the incomplete surface coverage by Al（OH）3. Fig.3（b） shows that in the Fe-Al （hydr）oxide nanocomposite， some of the Fe3O4 and Al（OH）3 characteristic diffraction peaks appeared at the same time （2q=13.8°， 28.5° from Al（OH）3 and 2q=35.6°， 57.1°， 62.8° from Fe3O4）. The Fe-Al （hydr）-oxide nano-/micro-particles are completely mixed and grown with each other and well dispersed by the wheat straw template， making a close relative intensity in all the characteristic peaks.
The magnetic hysteresis loops of Fe3O4 nanocomposite and Fe-Al （hydr）oxide nanocomposites with Fe to Al mole ratios of 4：1 and 1：4 are shown in Fig.4. Fe3O4 nanocomposite shows a typical superparamagnetic behavior[48] with the saturated magnetization of 46.13 emu/g， as shown in Fig.4（a）. With the mixing of Al（OH）3 particles， the Fe-Al （hydr）-oxide nanocomposite can maintain a certain magnetism when the mole ratio of Fe to Al is 4：1， with the saturated magnetization of 11.24 emu/g as shown in Fig.4（b）. When the Al composition is much higher， with an Fe to Al mole ratio of 1：4， the magnetism decreases， with a saturated magnetization of 0.23 emu/g as shown in Fig.4（c）. The magnetism of the Fe-Al （hydr）-oxide nanocomposite could be tuned by using different magnetic compositions of Fe3O4， and the nanocomposite material could be possibly separated under an external magnetic field as shown in the inset of Fig.4. 3.2 Fluoride adsorption behavior of the nanocomposite material
As shown in Fig.5， the adsorption capacity （qe） increases with the initial fluoride anion concentration （C0） in all Fe-Al （hydr）oxide nanocomposites with different mole ratios of Fe to Al. The Fe3O4 nanocomposite has no effect on F- adsorption， while the Fe-Al （hydr）oxide nanocomposite with an Fe to Al mole ratio of 4：1 has an obvious adsorption capacity of the fluoride anion， indicating that growing the mixed Fe-Al （hydr）oxides onto wheat straw with only a small amount of Al composition can add the advantage of fluoride adsorption.
The fluoride adsorption capacity qe increases as the Fe to Al mole ratio changes from 4：1 to 1：1， which is reasonably due to the good F- adsorption capacity of Al（OH）3. With further increase of Al composition in Fe-Al （hydr）oxides （Fe to Al mole ratio changes from 1：1 to 1：4）， the adsorption capacity qe of fluoride anion decreases， which is possibly due to the aggregation of the particles and the decrease of dispersing and loading effects through the wheat straw. As shown in Fig.5， it seems that the adsorption capacity of F- is most obvious when the mole ratio of Fe to Al is approximately 1：1 in the Fe-Al （hydr）oxide nanocomposite， and it is even higher than that of Al（OH）3 nanocomposite.
To understand the adsorption mechanism of the Fe-Al （hydr）oxide nanocomposite， we study the adsorption isotherms. In this study， the equilibrium condition was kept at room temperature. There are three main models of adsorption isotherms， which are the Langmuir model （monolayer adsorption， adsorption energy on the surface is uniform） expressed as Eq.（2）， Freundlich model （multilayer adsorption， adsorption energy decreases with surface coverage） expressed as Eq.（3）， and Temkin model （multilayer adsorption， the decrease of adsorption energy with surface coverage is linear rather than logarithmic） expressed as Eq.（4）.
In Eq.（2）， Ce （mg/L） is the equilibrium concentration and qe （mg/g） is the adsorption capacity. Q0 （mg/g） and b （L/mg） are the maximum adsorption capacity and adsorption energy related to Langmuir constants， respectively.
In Eq.（3）， Kf （mg/g） and n are the adsorption capacity and adsorption intensity of Freundlich constants， respectively.
In Eq.（4）， R is the gas constant 8.314 J/（mol·K） and T is the absolute temperature 298.15 K. AT （L/mg） and bT （J/mol） are the constants of the Temkin model.
The adsorption isotherms are studied by the three models， and all the fitted parameters are listed in Table 1. The fitted adsorption isotherms follow the Langmuir model better than the other two models， with a higher value of fitting correlation coefficient r2. It indicates that the process performed by nanocomposite material is much more like a monolayer adsorption. As shown in Fig.6， the adsorption isotherms of the Fe-Al （hydr）-oxide nanocomposite with different Fe to Al mole ratios for F- adsorption are well fitted by the Langmuir model. The constant Q0 of the Langmuir model represents the maximum adsorption capacity of the adsorbent. As presented in Fig.7 and Table 1， with the increase of Al composition， the value of Q0 increases first and reaches a maximum when the mole ratio of Fe to Al is approximately 1：1， and then decreases with further increase of Al composition. Generally， it is expected that the adsorption capacity of binary matrix composites would be between the two neat compositions. For Fe-Al （hydr）oxide nanocomposite， because Fe3O4 has no adsorption effect on fluoride anion， it is expected that with the increase of Fe composition in Fe-Al （hydr）-oxides， the value of Q0 would monotonically decrease. However， the results indicate that the adsorption capacity is unexpectedly higher with the intermediate Fe to Al mole ratios. In particular， the value of Q0 is obviously higher when the mole ratio of Fe to Al is 1：1， as shown in Fig.7. Such a synergistic adsorption of F- is found with the mixed Fe-Al （hydr）oxide nanocomposite， which has a higher value of Q0 than both Fe3O4 nanocomposite and Al（OH）3 nanocomposite. This could lead to the enhancement of adsorption capacity particularly when the Fe to Al mole ratio is 1：1.
To further determine the degree of the synergistic effect， we investigate the departure between the experimental and calculated values of （Q0-Q0（cal））. In Fe-Al （hydr）oxide nanocomposite， the adsorption mechanism is monolayer adsorption and Al（OH）3 is the only effective composition for F- adsorption. If the number of active sites for adsorption is proportional to the number of adsorbent molecules， under the same adsorption condition， the value of （C0-Ce） could be proportional to the mole number of the adsorbent. From this ideal condition， we can calculate Ce and qe from the adsorption results of Al（OH）3 nanocomposite and the mole number of Al（OH）3 in different samples. Then， we can obtain Q0（cal） from the fitting of the Langmuir model， and the value of Q0（cal） monotonically decreases with the decrease of Al composition in the material， as shown in Fig.7.
The value of （Q0-Q0（cal）） could reflect the difference between the maximum adsorption capacity in the experimental condition and that in the ideal calculation condition without considering other special effects. In Fig.7， the values of Q0 are higher than the corresponding Q0（cal） across all the materials with different Fe to Al mole ratios， which clearly indicates that there is a synergistic effect for F- adsorption by the Fe-Al （hydr） oxide nanocomposite. Moreover， the departure of （Q0-Q0（cal）） reaches a maximum near a 1：1 mole ratio of Fe to Al. Therefore， it is concluded that the maximum adsorption capacity of the nanocomposite material is enhanced by the synergistic effect with the different Fe to Al mole ratios， and with an approximately 1：1 mole ratio of Fe to Al， the degree of this unusual effect is most significant， which is possibly due to the good adsorption of fluoride by the mixed nano-/micro-particles well dispersed through the wheat straw. 3.3 Arsenic adsorption behavior of the nanocomposite material
The effect of the initial concentration and mole ratio of Fe to Al in Fe-Al （hydr）oxide nanocomposite on the adsorption capacity qe of arsenic （As（Ⅲ） and As（Ⅴ）） is shown in Fig.8. Both Fe3O4 and Al（OH）3 nano-/micro-particles in Fe-Al （hydr）oxide nanocomposites have adsorption effect on arsenic. With different Fe to Al mole ratios of the Fe-Al （hydr）oxide nanocomposite， generally the adsorption capacity qe of both As（Ⅲ） and As（Ⅴ） increases from Fe3O4 nanocomposite and Al（OH）3 nanocomposite to the materials with mixed Fe and Al compositions. Then， the adsorption capacity further increases when the Fe and Al mole ratio is close to 1：1， as shown in Fig.8（a） and Fig.8（b）. From the adsorption of both As（Ⅲ） and As（Ⅴ）， it is found that in many cases the adsorption capacity of the materials with the mixed Fe-Al （hydr）oxides is higher than that with either Al（OH）3 nanocomposite or Fe3O4 nanocomposite.
The adsorption isotherms for As（Ⅲ） and As（Ⅴ） are fitted by the Langmuir model， Freundlich model， and Temkin model using Eq.（2）， Eq.（3）， and Eq.（4）， and the obtained parameters are listed in Table 2 and Table 3. The results show that both As（Ⅲ） and As（Ⅴ） adsorption isotherms are well fitted by the Langmuir model with a higher fitting correlation coefficient （r2） than that of the other models， as shown in Fig.9（a） and Fig.9（b）. This indicates that the adsorption process prefers the monolayer adsorption by Fe-Al （hydr）oxide nanocomposite.
The maximum adsorption capacity Q0 from the fitting of the Langmuir model for arsenic （As（Ⅲ） and As（Ⅴ）） adsorption is shown in Fig.10. Regarding the effective compositions of Fe3O4 and Al（OH）3 for arsenic adsorption， Fe3O4 nanocomposite has a relatively higher value of Q0 for As（Ⅲ）， while Al（OH）3 nanocomposite has a higher value of Q0 for As（Ⅴ）. The value of the maximum adsorption capacity （Q0） for both As（Ⅲ） and As（Ⅴ） increases first and reaches a maximum when the Fe to Al mole ratio is 1：1， and then decreases with a further increase of Al composition， as shown in Fig.10（a） and Fig.10（b） （listed in Table 2 and Table 3）. Most values of Q0 for the materials with intermediate Fe to Al mole ratios of Fe-Al （hydr）oxides are unexpectedly higher than that for the material only loaded with Fe3O4 or Al（OH）3. Once again， the synergistic adsorption is clearly found from the adsorption of both As（Ⅲ） and As（Ⅴ）， with an obvious enhancement of adsorption capacity. This unusual behavior is most obvious with an Fe to Al mole ratio of 1：1， which results in a higher value of Q0 than that of all other nanocomposites. To further study this synergistic effect for arsenic adsorption， we calculate the value of Q0（cal） based on the adsorption results of Fe3O4 nanocomposite and Al（OH）3 nanocomposite， with the ideal addition for the adsorption capacity， and the mole number of Fe3O4 and Al（OH）3 in different samples， which is obtained from the fitting of the Langmuir model. The degree of this synergistic effect for arsenic adsorption could be determined by （Q0-Q0（cal））. As shown in Fig.10， all the values of the departure of （Q0-Q0（cal）） are positive for both the adsorption of As（Ⅲ） and As（Ⅴ）， and the value becomes a maximum when the Fe and Al mole ratio is approximately 1：1. The synergistic effect of Fe-Al （hydr）oxide nanocomposite for arsenic adsorption is obtained through all different Fe to Al mole ratios， and this unusual effect is most obvious when the Fe to Al mole ratio is approximately 1：1. This is possibly due to the good adsorption of arsenic by the mixed Fe and Al compositions and the good dispersity of nano-/micro-particles through the wheat straw.
Meanwhile， the maximum adsorption capacity （Q0） of the Fe-Al （hydr）oxide nanocomposite （n（Fe）：n（Al）=1：1） for fluoride and arsenic adsorption was compared with many similar studies， as listed in Table 4. Enhanced by the synergistic effect， both fluoride （F-） and arsenic （As（Ⅲ） and As（Ⅴ）） can be efficiently adsorbed by Fe-Al （hydr）oxide nanocomposite with competitive adsorption capacities. Moreover， the material could also maintain the advantage of magnetic separation. Besides， Fe-Al （hydr）oxide nanocomposite is easily obtained by a one-step synthesis with low-cost compounds and materials.
4 Conclusions
A biomass-based （wheat straw） Fe-Al （hydr）oxide nano-/micro-particle nanocomposite was obtained by a convenient， one-step synthesis. Favoring a monolayer adsorption process， the adsorption isotherms of the Fe-Al （hydr）oxide nanocomposite were well fitted by the Langmuir model. A synergistic adsorption was clearly found from both the adsorption of fluoride and arsenic with the enhancement of adsorption capacity， particularly when the Fe to Al mole ratio was 1：1. The synergistic effect determined by the departure of the experimental （Q0） and calculated （Q0（cal）） showed that this unusual effect was most pronounced with a 1：1 mole ratio of Fe and Al， which was possibly caused by the good adsorption of the mixed Fe-Al （hydr）oxides and the good dispersity of nano-/micro-particles by the wheat straw template. This Fe-Al （hydr）oxide nanocomposite could be a high-efficiency， low-cost adsorbent for the removal of fluoride （F-） and arsenic （As（Ⅲ） and As（Ⅴ）） from water with a magnetic separation. Acknowledgments
This study was supported by the National Natural Science Foundation of China （No. 51472253）， National Key Project of Research and Development Plan （2016YFC1402500） and Chinese Academy of Sciences Visiting Professorships.
References
[1] Schwarzenbach R P， Escher B I， Fenner K， et al. The challenge of micropollutants in aquatic systems[J]. Science， 2006， 313 （5790）： 1072-1077.
[2] Qiu J. China to spend billions cleaning up groundwater[J]. Science， 2011， 334 （6057）： 745-745.
[3] Banerjee K， Amy G L， Prevost M， et al. Kinetic and thermodynamic aspects of adsorption of arsenic onto granular ferric hydroxide （GFH）[J]. Water Res， 2008， 42 （13）： 3371-3378.
[4] An B， Liang Q， Zhao D. Removal of arsenic （V） from spent ion exchange brine using a new class of starch-bridged magnetite nanoparticles[J]. Water Res， 2011， 45（5）： 1961-1972.
[5] Tian Y， Wu M， Liu R， et al. Modified native cellulose fibers—A novel efficient adsorbent for both fluoride and arsenic[J]. J Hazard Mater， 2011， 185 （1）： 93-100.
[6] Wu Z， Li W， Webley P A， et al. General and controllable synthesis of novel mesoporous magnetic iron oxide@carbon encapsulates for efficient arsenic removal[J]. Adv Mater， 2012， 24 （4）： 485-491.
[7] Baker J L， Sudarsan N， Weinberg Z， et al. Widespread genetic switches and toxicity resistance proteins for fluoride[J]. Science， 2012， 335 （6065）： 233-235.
[8] Zhang Q， Du Q， Jiao T， et al. Rationally designed porous polystyrene encapsulated zirconium phosphate nanocomposite for highly efficient fluoride uptake in waters[J]. Sci Rep， 2013， DOI： 10.1038/srep02551.
[9] Rodríguez-Lado L， Sun G， Berg M， et al. Groundwater arsenic contamination throughout China[J]. Science， 2013， 341（6148）： 866-868.
[10] Jing C， Cui J， Huang Y， et al. Fabrication， characterization， and application of a composite adsorbent for simultaneous removal of arsenic and fluoride[J]. ACS Appl Mater Interfaces， 2012， 4（2）： 714-720.
[11] Liu M， Wang Y， Chen L， et al. Mg（OH）2 Supported Nanoscale Zero Valent Iron Enhancing the Removal of Pb（II） from Aqueous Solution[J]. ACS Appl Mater Interfaces， 2015， 7（15）： 7961-7969.
[12] Zularisam A， Ismail A， Salim R. Behaviours of natural organic matter in membrane filtration for surface water treatment—a review[J]. Desalination， 2006， 194（1）： 211-231.
[13] Hollender J， Zimmermann S G， Koepke S， et al. Elimination of organic micropollutants in a municipal wastewater treatment plant upgraded with a full-scale post-ozonation followed by sand filtration[J]. Environ Sci Technol， 2009， 43 （20）： 7862-7869. [14] Ndiaye P， Moulin P， Dominguez L， et al. Removal of fluoride from electronic industrial effluentby RO membrane separation[J]. Desalination， 2005， 173（1）： 25-32.
[15] Kota A K， Kwon G， Choi W， et al. Hygro-responsive membranes for effective oil-water separation[J]. Nat Commun， 2012， DOI： 10.1038/ncomms2027.
[16] Lin S H， Peng C F. Continuous treatment of textile wastewater by combined coagulation， electrochemical oxidation and activated sludge[J]. Water Res， 1996， 30（3）： 587-592.
[17] Verma A K， Dash R R， Bhunia P. A review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters[J]. J Environ Manage， 2012， 93（1）： 154-168.
[18] Jorgensen T， Weatherley L. Ammonia removal from wastewater by ion exchange in the presence of organic contaminants[J]. Water Res， 2003， 37（8）： 1723-1728.
[19] Raghu S， Basha C A. Chemical or electrochemical techniques， followed by ion exchange， for recycle of textile dye wastewater[J]. J Hazard Mater， 2007， 149（2）： 324-330.
[20] Huber M M， G?bel A， Joss A， et al. Oxidation of pharmaceuticals during ozonation of municipal wastewater effluents： a pilot study[J]. Environ Sci Technol， 2005， 39（11）： 4290-4299.
[21] Chatterjee D， Rothbart S， Eldik R. Substrate versus oxidant activation in Ru Ⅲ （edta） catalyzed dye degradation[J]. RSC Adv， 2013， 3（11）： 3606-3610.
[22] Fan L， Ni J， Wu Y， et al. Treatment of bromoamine acid wastewater using combined process of micro-electrolysis and biological aerobic filter[J]. J Hazard Mater， 2009， 162（2）： 1204-1210.
[23] Reyter D， Bélanger D， Roué L. Nitrate removal by a paired electrolysis on copper and Ti/IrO2 coupled electrodes-influence of the anode/cathode surface area ratio[J]. Water Res， 2010， 44（6）： 1918-1926.
[24] Bhatnagar A， Kumar E， Sillanp?? M. Fluoride removal from water by adsorption—a review[J]. Chem Eng J， 2011， 171（3）： 811-840.
[25] Cao C Y， Qu J， Yan W S， et al. Low-cost synthesis of flowerlike a-Fe2O3 nanostructures for heavy metal ion removal： adsorption property and mechanism[J]. Langmuir， 2012， 28（9）： 4573-4579.
[26] Yu X， Tong S， Ge M， et al. One-step synthesis of magnetic composites of cellulose@ iron oxide nanoparticles for arsenic removal[J]. J Mater Chem A， 2013， DOI： 10.1039/c2ta00315e.
[27] Li W， Cao C Y， Wu L Y， et al. Superb fluoride and arsenic removal performance of highly ordered mesoporous aluminas[J]. J Hazard Mater， 2011， 198： 143-150. [28] Zhong L S， Hu J S， Liang H P， et al. Self-Assembled 3D flowerlike iron oxide nanostructures and their application in water treatment[J]. Adv Mater， 2006， 18（18）： 2426-2431.
[29] Cao C Y， Qu J， Wei F， et al. Superb adsorption capacity and mechanism of flowerlike magnesium oxide nanostructures for lead and cadmium ions[J]. ACS Appl Mater Interfaces， 2012， 4（8）： 4283-4287.
[30] Chen L， He S， He B Y， et al. Synthesis of iron-doped titanium oxide nanoadsorbent and its adsorption characteristics for fluoride in drinking water[J]. Ind Eng Chem Res， 2012， 51（40）： 13150-13156.
[31] Ma R， Levard C， Judy J D， et al. Fate of zinc oxide and silver nanoparticles in a pilot wastewater treatment plant and in processed biosolids[J]. Environ Sci Technol， 2013， 48 （1）： 104-112.
[32] Chandra V， Park J， Chun Y， et al. Water-dispersible magnetite-reduced graphene oxide composites for arsenic removal[J]. ACS Nano， 2010， 4（7）： 3979-3986.
[33] Lu W， Gao S， Wang J. One-pot synthesis of Ag/ZnO self-assembled 3D hollow microspheres with enhanced photocatalytic performance[J]. J Phys Chem C， 2008， 112（43）： 16792-16800.
[34] Wang Y， Zou B， Gao T， et al. Synthesis of orange-like Fe3O4/PPy composite microspheres and their excellent Cr （VI） ion removal properties[J]. J Mater Chem， 2012， 22（18）： 9034-9040.
[35] Kang D， Tong S， Yu X， et al. Template-free synthesis of 3D hierarchical amorphous aluminum oxide microspheres with broccoli-like structure and their application in fluoride removal[J]. RSC Adv， 2015， 5（25）： 19159-19165.
[36] Kang D， Yu X， Ge M， et al. One-step fabrication and characterization of hierarchical MgFe2O4 microspheres and their application for lead removal[J]. Micropor Mesopor Mater， 2015， 207： 170-178.
[37] Mohan D， Singh K P. Single-and multi-component adsorption of cadmium and zinc using activated carbon derived from bagasse—an agricultural waste[J]. Water Res， 2002， 36（9）： 2304-2318.
[38] Pan B， Xing B. Adsorption mechanisms of organic chemicals on carbon nanotubes[J]. Environ Sci Technol， 2008， 42（24）： 9005-9013.
[39] Yang X， Zhen M， Li G， et al. Preparation of Pd-decorated fullerenols on carbon nanotubes with excellent electrocatalytic properties in alkaline media[J]. J Mater Chem A， 2013， DOI：10.1039/c3ta11907f.
[40] Kang D， Yu X， Tong S， et al. Performance and mechanism of Mg/Fe layered double hydroxides for fluoride and arsenate removal from aqueous solution[J]. Chem Eng J， 2013： 228， 731-740. [41] Xiong R， Wang Y， Zhang X， et al. Facile synthesis of magnetic nanocomposites of cellulose@ ultrasmall iron oxide nanoparticles for water treatment[J]. RSC Adv， 2014， 4（43）： 22632-22641.
[42] Yu X， Kang D， Hu Y， et al. One-pot synthesis of porous magnetic cellulose beads for the removal of metal ions[J]. RSC Adv， 2014， 4（59）： 31362-31369.
[43] Tian Y， Wu M， Lin X， et al. Synthesis of magnetic wheat straw for arsenic adsorption[J]. J Hazard Mater， 2011， 193： 10-16.
[44] Liu R， Yu H， Huang Y. Structure and morphology of cellulose in wheat straw[J]. Cellulose， 2005， 12（1）： 25-34.
[45] Yu H， Liu R， Shen D， et al. Study on morphology and orientation of cellulose in the vascular bundle of wheat straw[J]. Polymer， 2005， 46（15）： 5689-5694.
[46] Cao M S， Yang J， Song W L， et al. Ferroferric oxide/multiwalled carbon nanotube vs polyaniline/ferroferric oxide/multiwalled carbon nanotube multiheterostructures for highly effective microwave absorption[J]. ACS Appl Mater Interfaces， 2012， 4（12）： 6949-6956.
[47] Tanada S， Kabayama M， Kawasaki N， et al. Removal of phosphate by aluminum oxide hydroxide[J]. J Coll Interf Sci， 2003， 257（1）： 135-140.
[48] Deng Y， Qi D， Deng C， et al. Superparamagnetic high-magnetization microspheres with an Fe3O4@SiO2 core and perpendicularly aligned mesoporous SiO2 shell for removal of microcystins[J]. J Am Chem Soc， 2008， 130（1）： 28-29.
Keywords： adsorption; synergistic effect; nanocomposite; biomass; fluoride; arsenic
1 Introduction
Anion pollution in water is a worldwide problem harming human health and natural ecosystems[1-6]. Fluoride is a necessary element for humans; however， the excessive ingestion of fluoride may cause many chronic diseases[7-8]. Arsenic， which mainly exists as arsenite and arsenate in nature， can seriously damage human systems and increase cancer risk[9]. In many areas， water resources continuously deteriorate and are possibly contaminated by both fluoride and arsenic[10]. Therefore， the removal of these harmful anions from water is critical.
Many techniques have been developed for the purification of water， such as precipitation[11]， filtration[12-13]， membrane separation[14-15]， coagulation[16-17]， ion exchange[18-19]， oxidation[20-21]， electrolysis[22-23]， and adsorption[24-26]. Among them， adsorption is one of the most important methods because of its wide applicability， easy operation， high efficiency， and low cost. Nanoparticles have attracted significant interest for the adsorption and removal of pollutants from water owing to their high efficiency based on large surface areas. Many nanoparticle adsorbents composed of alumina[27]， iron oxide[28]， magnesium oxide[29]， titanium oxide[30]， and zinc oxide[31] have been applied to water treatment. A general problem of nanoparticle adsorbents is their aggregation， which leads to the decrease in the number of active sites[6]. In addition， the residue of nanoparticles in water could cause a secondary pollution. To avoid the aggregation of nanoparticles， template materials to disperse them， including graphene oxide[32]， microspheres[33-36]， carbon[37-39]， layered double hydroxides[40]， and cellulose[41-42]， have been extensively studied. Wheat straw is an eco-friendly and low-cost biomass-based material available in large amounts， which is mainly composed of cellulose， hemicelluloses， and lignin with numerous hydroxyl groups. Metal ions could be adsorbed by the hydroxyl groups on the surface of wheat straw and grown to nano-/micro-particles[43]. With a vascular bundle structure and hydroxyl groups on its surface[44-45]， wheat straw could be a suitable matrix to disperse nanoparticles. A previous study showed that the aggregation of magnetic nano-/micro-particles can be efficiently avoided by growing and dispersing them onto wheat straw， and the material can be used for adsorption and easily separated from water[43].
In this study， Fe-Al （hydr）oxide nano-/micro-particles were simultaneously grown onto wheat straw through a co-precipitation method. A synergistic adsorption by the mixed Fe-Al （hydr）oxide nano-/micro-particles dispersed onto wheat straw was unexpectedly found with the enhancement of adsorption capacity. The synergistic effect was determined by the departure between the experimental and calculated maximum adsorption capacity fitted from Langmuir adsorption isotherms. The synergistic effect for the adsorption was most pronounced when the Fe and Al mole ratio was approximately 1：1. This material could be used for the removal of fluoride and arsenic and be separated from water under a magnetic field.
2 Experimental
2.1 Materials
Wheat straw was obtained from the countryside of China. After being cleaned and dried， the wheat straw was mechanically ground into fragments by a miniature plant grinding machine （FZ102， Tianjin City Taisite Instrument Co.， Ltd）. The specific surface area of the wheat straw fragments is 1.91 m2/g. Ferric chloride （FeCl3·6H2O）， ferrous sulfate （FeSO4·7H2O）， aluminum nitrate （Al（NO3）3·9H2O）， and ammonia water （NH3·H2O， 25 wt%） were used for the synthesis of Fe-Al （hydr）oxide nano-/micro-particles. Sodium fluoride （NaF）， sodium arsenite （NaAsO2）， and disodium hydrogen arsenate （Na2HAsO4·7H2O） were used for the test solutions. All the chemicals were purchased from Beijing Chemical Works （Beijing， China） with an analytical grade and used without further purification.
2.2 Synthesis of biomass-based Fe-Al （hydr）oxide nanocomposite The biomass-based Fe-Al （hydr）oxide nanocomposite was synthesized through a co-precipitation method. Approximately 0.8 g of wheat straw fragments were suspended in a 50-mL solution with FeCl3， FeSO4 （the mole ratio of Fe3+ to Fe2+ is 2：1）， and Al（NO3）3. The total ion concentration of the solution was 0.5 mol/L， with mole ratios of total Fe to Al set as 4：1， 3：2， 1：1， 2：3， and 1：4. Under nitrogen atmosphere， 10 mL of ammonia water （25 wt%） was added dropwise into the solution with vigorous stirring at 70℃ for 4 h. The synthesized materials were separated and washed by deionized water several times， and then dried for use. Wheat straw-based nanocomposite loaded only with Fe3O4 or Al（OH）3 was also synthesized for comparison （the products named as Fe3O4 nanocomposite and Al（OH）3 nanocomposite， respectively）.
2.3 Characterization of biomass-based Fe-Al （hydr）-oxide nanocomposite
The specific surface area of the wheat straw fragments was measured through nitrogen adsorption by the BET method using an automated gas sorption analyzer （Quadrasorb SI-MP）. The morphologies of the synthesized biomass-based Fe-Al （hydr）oxide nanocomposites were observed by a scanning electron microscope （SEM） （S-4800， Hitachi） equipped with energy-dispersive X-ray spectroscopy （EDS）. X-ray diffraction （XRD） was studied by using a D8 Focus （Bruker） for crystal characterization. The magnetic properties were characterized using a vibrating sample magnetometer （SQUID-VSM， Quantum Design）.
2.4 Adsorption experiments
Adsorption experiments were conducted in an aqueous solution with different initial concentrations of ions under stirring at room temperature for 24 h. The adsorbent dose was 1 g/L for all the experiments. The concentrations of fluoride ion were measured by a fluoride analyzer （PXSJ-216， INESA）. Arsenic was quantified by an inductive coupled plasma emission spectrometer （ICP） （710-OES， Varian）. The adsorption capacity qe （mg/g） was calculated from the following equation （Eq.（1））：
Where C0 （mg/L） and Ce （mg/L） represent the initial and equilibrium concentration of adsorbate in the solution， respectively. V （L） is the solution volume， and m （g） is the adsorbent mass.
3 Results and discussion
3.1 Characterization of the nanocomposite material
The surface morphologies of wheat straw fragments and the nano-/micro-particle-loaded nanocomposites were studied by SEM， as shown in Fig.1. The surface of the original wheat straw fragments is very smooth （Fig.1（a））. Growth of Fe-Al （hydr）oxides （n（Fe）：n（Al）=1：1） results in the flocculent deposition of well dispersed nano-/micro-particles as shown in Fig.1（b）， indicating the suitability of wheat straw as a template. Fig.1（c） and Fig.1（d） show that Fe3O4 nanoparticles are well dispersed by the wheat straw， while the Al（OH）3 particles have almost grown to micrometer size and are coarsely distributed. As shown in Fig.2（a）～Fig.2（c）， the mixed Fe-Al （hydr）oxide nano-/micro-particles （n（Fe）：n（Al）=1：1） could well grow to coral-like floccules， and both Fe3O4 and Al（OH）3 compositions are uniformly distributed on the surface of the wheat straw matrix investigated by EDS elemental mapping. The co-precipitation formation process of Fe-Al （hydr）oxides onto wheat straw is likely to be effective in avoiding the aggregation of particles， as shown in Fig.2（d）.
The XRD patterns of wheat straw fragmets， Fe-Al （hydr）oxide nanocomposite （n（Fe）：n（Al）=1：1）， Fe3O4 nanocomposite， and Al（OH）3 nanocomposite are shown in Fig.3. The diffraction peaks at 2q=16.2° and 22.2° are those of native cellulose I （Fig.3（a））[44]. Loading with Fe-Al （hydr）oxide nano-/micro-particles causes the appearance of many other diffraction peaks （Fig.3（b））， nearly hiding the cellulose peaks. That is because the full growth of Fe-Al （hydr）oxides coats the surface of the wheat straw. In Fig.3（c） and Fig.3（d）， the characteristic diffraction peaks of Fe3O4 and Al（OH）3 are found. The Fe3O4 diffraction peaks are at 2q=30.4°， 35.6°， 43.3°， 57.1°， 62.8°[46] and the typical Al（OH）3 characteristic diffraction peaks are at 2q=13.8°， 28.5°， 38.4°， 49.1°[47]. The diffraction peak at 2q=22.2° of cellulose is relatively more obvious in Al（OH）3 nanocomposite， which is possibly due to the incomplete surface coverage by Al（OH）3. Fig.3（b） shows that in the Fe-Al （hydr）oxide nanocomposite， some of the Fe3O4 and Al（OH）3 characteristic diffraction peaks appeared at the same time （2q=13.8°， 28.5° from Al（OH）3 and 2q=35.6°， 57.1°， 62.8° from Fe3O4）. The Fe-Al （hydr）-oxide nano-/micro-particles are completely mixed and grown with each other and well dispersed by the wheat straw template， making a close relative intensity in all the characteristic peaks.
The magnetic hysteresis loops of Fe3O4 nanocomposite and Fe-Al （hydr）oxide nanocomposites with Fe to Al mole ratios of 4：1 and 1：4 are shown in Fig.4. Fe3O4 nanocomposite shows a typical superparamagnetic behavior[48] with the saturated magnetization of 46.13 emu/g， as shown in Fig.4（a）. With the mixing of Al（OH）3 particles， the Fe-Al （hydr）-oxide nanocomposite can maintain a certain magnetism when the mole ratio of Fe to Al is 4：1， with the saturated magnetization of 11.24 emu/g as shown in Fig.4（b）. When the Al composition is much higher， with an Fe to Al mole ratio of 1：4， the magnetism decreases， with a saturated magnetization of 0.23 emu/g as shown in Fig.4（c）. The magnetism of the Fe-Al （hydr）-oxide nanocomposite could be tuned by using different magnetic compositions of Fe3O4， and the nanocomposite material could be possibly separated under an external magnetic field as shown in the inset of Fig.4. 3.2 Fluoride adsorption behavior of the nanocomposite material
As shown in Fig.5， the adsorption capacity （qe） increases with the initial fluoride anion concentration （C0） in all Fe-Al （hydr）oxide nanocomposites with different mole ratios of Fe to Al. The Fe3O4 nanocomposite has no effect on F- adsorption， while the Fe-Al （hydr）oxide nanocomposite with an Fe to Al mole ratio of 4：1 has an obvious adsorption capacity of the fluoride anion， indicating that growing the mixed Fe-Al （hydr）oxides onto wheat straw with only a small amount of Al composition can add the advantage of fluoride adsorption.
The fluoride adsorption capacity qe increases as the Fe to Al mole ratio changes from 4：1 to 1：1， which is reasonably due to the good F- adsorption capacity of Al（OH）3. With further increase of Al composition in Fe-Al （hydr）oxides （Fe to Al mole ratio changes from 1：1 to 1：4）， the adsorption capacity qe of fluoride anion decreases， which is possibly due to the aggregation of the particles and the decrease of dispersing and loading effects through the wheat straw. As shown in Fig.5， it seems that the adsorption capacity of F- is most obvious when the mole ratio of Fe to Al is approximately 1：1 in the Fe-Al （hydr）oxide nanocomposite， and it is even higher than that of Al（OH）3 nanocomposite.
To understand the adsorption mechanism of the Fe-Al （hydr）oxide nanocomposite， we study the adsorption isotherms. In this study， the equilibrium condition was kept at room temperature. There are three main models of adsorption isotherms， which are the Langmuir model （monolayer adsorption， adsorption energy on the surface is uniform） expressed as Eq.（2）， Freundlich model （multilayer adsorption， adsorption energy decreases with surface coverage） expressed as Eq.（3）， and Temkin model （multilayer adsorption， the decrease of adsorption energy with surface coverage is linear rather than logarithmic） expressed as Eq.（4）.
In Eq.（2）， Ce （mg/L） is the equilibrium concentration and qe （mg/g） is the adsorption capacity. Q0 （mg/g） and b （L/mg） are the maximum adsorption capacity and adsorption energy related to Langmuir constants， respectively.
In Eq.（3）， Kf （mg/g） and n are the adsorption capacity and adsorption intensity of Freundlich constants， respectively.
In Eq.（4）， R is the gas constant 8.314 J/（mol·K） and T is the absolute temperature 298.15 K. AT （L/mg） and bT （J/mol） are the constants of the Temkin model.
The adsorption isotherms are studied by the three models， and all the fitted parameters are listed in Table 1. The fitted adsorption isotherms follow the Langmuir model better than the other two models， with a higher value of fitting correlation coefficient r2. It indicates that the process performed by nanocomposite material is much more like a monolayer adsorption. As shown in Fig.6， the adsorption isotherms of the Fe-Al （hydr）-oxide nanocomposite with different Fe to Al mole ratios for F- adsorption are well fitted by the Langmuir model. The constant Q0 of the Langmuir model represents the maximum adsorption capacity of the adsorbent. As presented in Fig.7 and Table 1， with the increase of Al composition， the value of Q0 increases first and reaches a maximum when the mole ratio of Fe to Al is approximately 1：1， and then decreases with further increase of Al composition. Generally， it is expected that the adsorption capacity of binary matrix composites would be between the two neat compositions. For Fe-Al （hydr）oxide nanocomposite， because Fe3O4 has no adsorption effect on fluoride anion， it is expected that with the increase of Fe composition in Fe-Al （hydr）-oxides， the value of Q0 would monotonically decrease. However， the results indicate that the adsorption capacity is unexpectedly higher with the intermediate Fe to Al mole ratios. In particular， the value of Q0 is obviously higher when the mole ratio of Fe to Al is 1：1， as shown in Fig.7. Such a synergistic adsorption of F- is found with the mixed Fe-Al （hydr）oxide nanocomposite， which has a higher value of Q0 than both Fe3O4 nanocomposite and Al（OH）3 nanocomposite. This could lead to the enhancement of adsorption capacity particularly when the Fe to Al mole ratio is 1：1.
To further determine the degree of the synergistic effect， we investigate the departure between the experimental and calculated values of （Q0-Q0（cal））. In Fe-Al （hydr）oxide nanocomposite， the adsorption mechanism is monolayer adsorption and Al（OH）3 is the only effective composition for F- adsorption. If the number of active sites for adsorption is proportional to the number of adsorbent molecules， under the same adsorption condition， the value of （C0-Ce） could be proportional to the mole number of the adsorbent. From this ideal condition， we can calculate Ce and qe from the adsorption results of Al（OH）3 nanocomposite and the mole number of Al（OH）3 in different samples. Then， we can obtain Q0（cal） from the fitting of the Langmuir model， and the value of Q0（cal） monotonically decreases with the decrease of Al composition in the material， as shown in Fig.7.
The value of （Q0-Q0（cal）） could reflect the difference between the maximum adsorption capacity in the experimental condition and that in the ideal calculation condition without considering other special effects. In Fig.7， the values of Q0 are higher than the corresponding Q0（cal） across all the materials with different Fe to Al mole ratios， which clearly indicates that there is a synergistic effect for F- adsorption by the Fe-Al （hydr） oxide nanocomposite. Moreover， the departure of （Q0-Q0（cal）） reaches a maximum near a 1：1 mole ratio of Fe to Al. Therefore， it is concluded that the maximum adsorption capacity of the nanocomposite material is enhanced by the synergistic effect with the different Fe to Al mole ratios， and with an approximately 1：1 mole ratio of Fe to Al， the degree of this unusual effect is most significant， which is possibly due to the good adsorption of fluoride by the mixed nano-/micro-particles well dispersed through the wheat straw. 3.3 Arsenic adsorption behavior of the nanocomposite material
The effect of the initial concentration and mole ratio of Fe to Al in Fe-Al （hydr）oxide nanocomposite on the adsorption capacity qe of arsenic （As（Ⅲ） and As（Ⅴ）） is shown in Fig.8. Both Fe3O4 and Al（OH）3 nano-/micro-particles in Fe-Al （hydr）oxide nanocomposites have adsorption effect on arsenic. With different Fe to Al mole ratios of the Fe-Al （hydr）oxide nanocomposite， generally the adsorption capacity qe of both As（Ⅲ） and As（Ⅴ） increases from Fe3O4 nanocomposite and Al（OH）3 nanocomposite to the materials with mixed Fe and Al compositions. Then， the adsorption capacity further increases when the Fe and Al mole ratio is close to 1：1， as shown in Fig.8（a） and Fig.8（b）. From the adsorption of both As（Ⅲ） and As（Ⅴ）， it is found that in many cases the adsorption capacity of the materials with the mixed Fe-Al （hydr）oxides is higher than that with either Al（OH）3 nanocomposite or Fe3O4 nanocomposite.
The adsorption isotherms for As（Ⅲ） and As（Ⅴ） are fitted by the Langmuir model， Freundlich model， and Temkin model using Eq.（2）， Eq.（3）， and Eq.（4）， and the obtained parameters are listed in Table 2 and Table 3. The results show that both As（Ⅲ） and As（Ⅴ） adsorption isotherms are well fitted by the Langmuir model with a higher fitting correlation coefficient （r2） than that of the other models， as shown in Fig.9（a） and Fig.9（b）. This indicates that the adsorption process prefers the monolayer adsorption by Fe-Al （hydr）oxide nanocomposite.
The maximum adsorption capacity Q0 from the fitting of the Langmuir model for arsenic （As（Ⅲ） and As（Ⅴ）） adsorption is shown in Fig.10. Regarding the effective compositions of Fe3O4 and Al（OH）3 for arsenic adsorption， Fe3O4 nanocomposite has a relatively higher value of Q0 for As（Ⅲ）， while Al（OH）3 nanocomposite has a higher value of Q0 for As（Ⅴ）. The value of the maximum adsorption capacity （Q0） for both As（Ⅲ） and As（Ⅴ） increases first and reaches a maximum when the Fe to Al mole ratio is 1：1， and then decreases with a further increase of Al composition， as shown in Fig.10（a） and Fig.10（b） （listed in Table 2 and Table 3）. Most values of Q0 for the materials with intermediate Fe to Al mole ratios of Fe-Al （hydr）oxides are unexpectedly higher than that for the material only loaded with Fe3O4 or Al（OH）3. Once again， the synergistic adsorption is clearly found from the adsorption of both As（Ⅲ） and As（Ⅴ）， with an obvious enhancement of adsorption capacity. This unusual behavior is most obvious with an Fe to Al mole ratio of 1：1， which results in a higher value of Q0 than that of all other nanocomposites. To further study this synergistic effect for arsenic adsorption， we calculate the value of Q0（cal） based on the adsorption results of Fe3O4 nanocomposite and Al（OH）3 nanocomposite， with the ideal addition for the adsorption capacity， and the mole number of Fe3O4 and Al（OH）3 in different samples， which is obtained from the fitting of the Langmuir model. The degree of this synergistic effect for arsenic adsorption could be determined by （Q0-Q0（cal））. As shown in Fig.10， all the values of the departure of （Q0-Q0（cal）） are positive for both the adsorption of As（Ⅲ） and As（Ⅴ）， and the value becomes a maximum when the Fe and Al mole ratio is approximately 1：1. The synergistic effect of Fe-Al （hydr）oxide nanocomposite for arsenic adsorption is obtained through all different Fe to Al mole ratios， and this unusual effect is most obvious when the Fe to Al mole ratio is approximately 1：1. This is possibly due to the good adsorption of arsenic by the mixed Fe and Al compositions and the good dispersity of nano-/micro-particles through the wheat straw.
Meanwhile， the maximum adsorption capacity （Q0） of the Fe-Al （hydr）oxide nanocomposite （n（Fe）：n（Al）=1：1） for fluoride and arsenic adsorption was compared with many similar studies， as listed in Table 4. Enhanced by the synergistic effect， both fluoride （F-） and arsenic （As（Ⅲ） and As（Ⅴ）） can be efficiently adsorbed by Fe-Al （hydr）oxide nanocomposite with competitive adsorption capacities. Moreover， the material could also maintain the advantage of magnetic separation. Besides， Fe-Al （hydr）oxide nanocomposite is easily obtained by a one-step synthesis with low-cost compounds and materials.
4 Conclusions
A biomass-based （wheat straw） Fe-Al （hydr）oxide nano-/micro-particle nanocomposite was obtained by a convenient， one-step synthesis. Favoring a monolayer adsorption process， the adsorption isotherms of the Fe-Al （hydr）oxide nanocomposite were well fitted by the Langmuir model. A synergistic adsorption was clearly found from both the adsorption of fluoride and arsenic with the enhancement of adsorption capacity， particularly when the Fe to Al mole ratio was 1：1. The synergistic effect determined by the departure of the experimental （Q0） and calculated （Q0（cal）） showed that this unusual effect was most pronounced with a 1：1 mole ratio of Fe and Al， which was possibly caused by the good adsorption of the mixed Fe-Al （hydr）oxides and the good dispersity of nano-/micro-particles by the wheat straw template. This Fe-Al （hydr）oxide nanocomposite could be a high-efficiency， low-cost adsorbent for the removal of fluoride （F-） and arsenic （As（Ⅲ） and As（Ⅴ）） from water with a magnetic separation. Acknowledgments
This study was supported by the National Natural Science Foundation of China （No. 51472253）， National Key Project of Research and Development Plan （2016YFC1402500） and Chinese Academy of Sciences Visiting Professorships.
References
[1] Schwarzenbach R P， Escher B I， Fenner K， et al. The challenge of micropollutants in aquatic systems[J]. Science， 2006， 313 （5790）： 1072-1077.
[2] Qiu J. China to spend billions cleaning up groundwater[J]. Science， 2011， 334 （6057）： 745-745.
[3] Banerjee K， Amy G L， Prevost M， et al. Kinetic and thermodynamic aspects of adsorption of arsenic onto granular ferric hydroxide （GFH）[J]. Water Res， 2008， 42 （13）： 3371-3378.
[4] An B， Liang Q， Zhao D. Removal of arsenic （V） from spent ion exchange brine using a new class of starch-bridged magnetite nanoparticles[J]. Water Res， 2011， 45（5）： 1961-1972.
[5] Tian Y， Wu M， Liu R， et al. Modified native cellulose fibers—A novel efficient adsorbent for both fluoride and arsenic[J]. J Hazard Mater， 2011， 185 （1）： 93-100.
[6] Wu Z， Li W， Webley P A， et al. General and controllable synthesis of novel mesoporous magnetic iron oxide@carbon encapsulates for efficient arsenic removal[J]. Adv Mater， 2012， 24 （4）： 485-491.
[7] Baker J L， Sudarsan N， Weinberg Z， et al. Widespread genetic switches and toxicity resistance proteins for fluoride[J]. Science， 2012， 335 （6065）： 233-235.
[8] Zhang Q， Du Q， Jiao T， et al. Rationally designed porous polystyrene encapsulated zirconium phosphate nanocomposite for highly efficient fluoride uptake in waters[J]. Sci Rep， 2013， DOI： 10.1038/srep02551.
[9] Rodríguez-Lado L， Sun G， Berg M， et al. Groundwater arsenic contamination throughout China[J]. Science， 2013， 341（6148）： 866-868.
[10] Jing C， Cui J， Huang Y， et al. Fabrication， characterization， and application of a composite adsorbent for simultaneous removal of arsenic and fluoride[J]. ACS Appl Mater Interfaces， 2012， 4（2）： 714-720.
[11] Liu M， Wang Y， Chen L， et al. Mg（OH）2 Supported Nanoscale Zero Valent Iron Enhancing the Removal of Pb（II） from Aqueous Solution[J]. ACS Appl Mater Interfaces， 2015， 7（15）： 7961-7969.
[12] Zularisam A， Ismail A， Salim R. Behaviours of natural organic matter in membrane filtration for surface water treatment—a review[J]. Desalination， 2006， 194（1）： 211-231.
[13] Hollender J， Zimmermann S G， Koepke S， et al. Elimination of organic micropollutants in a municipal wastewater treatment plant upgraded with a full-scale post-ozonation followed by sand filtration[J]. Environ Sci Technol， 2009， 43 （20）： 7862-7869. [14] Ndiaye P， Moulin P， Dominguez L， et al. Removal of fluoride from electronic industrial effluentby RO membrane separation[J]. Desalination， 2005， 173（1）： 25-32.
[15] Kota A K， Kwon G， Choi W， et al. Hygro-responsive membranes for effective oil-water separation[J]. Nat Commun， 2012， DOI： 10.1038/ncomms2027.
[16] Lin S H， Peng C F. Continuous treatment of textile wastewater by combined coagulation， electrochemical oxidation and activated sludge[J]. Water Res， 1996， 30（3）： 587-592.
[17] Verma A K， Dash R R， Bhunia P. A review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters[J]. J Environ Manage， 2012， 93（1）： 154-168.
[18] Jorgensen T， Weatherley L. Ammonia removal from wastewater by ion exchange in the presence of organic contaminants[J]. Water Res， 2003， 37（8）： 1723-1728.
[19] Raghu S， Basha C A. Chemical or electrochemical techniques， followed by ion exchange， for recycle of textile dye wastewater[J]. J Hazard Mater， 2007， 149（2）： 324-330.
[20] Huber M M， G?bel A， Joss A， et al. Oxidation of pharmaceuticals during ozonation of municipal wastewater effluents： a pilot study[J]. Environ Sci Technol， 2005， 39（11）： 4290-4299.
[21] Chatterjee D， Rothbart S， Eldik R. Substrate versus oxidant activation in Ru Ⅲ （edta） catalyzed dye degradation[J]. RSC Adv， 2013， 3（11）： 3606-3610.
[22] Fan L， Ni J， Wu Y， et al. Treatment of bromoamine acid wastewater using combined process of micro-electrolysis and biological aerobic filter[J]. J Hazard Mater， 2009， 162（2）： 1204-1210.
[23] Reyter D， Bélanger D， Roué L. Nitrate removal by a paired electrolysis on copper and Ti/IrO2 coupled electrodes-influence of the anode/cathode surface area ratio[J]. Water Res， 2010， 44（6）： 1918-1926.
[24] Bhatnagar A， Kumar E， Sillanp?? M. Fluoride removal from water by adsorption—a review[J]. Chem Eng J， 2011， 171（3）： 811-840.
[25] Cao C Y， Qu J， Yan W S， et al. Low-cost synthesis of flowerlike a-Fe2O3 nanostructures for heavy metal ion removal： adsorption property and mechanism[J]. Langmuir， 2012， 28（9）： 4573-4579.
[26] Yu X， Tong S， Ge M， et al. One-step synthesis of magnetic composites of cellulose@ iron oxide nanoparticles for arsenic removal[J]. J Mater Chem A， 2013， DOI： 10.1039/c2ta00315e.
[27] Li W， Cao C Y， Wu L Y， et al. Superb fluoride and arsenic removal performance of highly ordered mesoporous aluminas[J]. J Hazard Mater， 2011， 198： 143-150. [28] Zhong L S， Hu J S， Liang H P， et al. Self-Assembled 3D flowerlike iron oxide nanostructures and their application in water treatment[J]. Adv Mater， 2006， 18（18）： 2426-2431.
[29] Cao C Y， Qu J， Wei F， et al. Superb adsorption capacity and mechanism of flowerlike magnesium oxide nanostructures for lead and cadmium ions[J]. ACS Appl Mater Interfaces， 2012， 4（8）： 4283-4287.
[30] Chen L， He S， He B Y， et al. Synthesis of iron-doped titanium oxide nanoadsorbent and its adsorption characteristics for fluoride in drinking water[J]. Ind Eng Chem Res， 2012， 51（40）： 13150-13156.
[31] Ma R， Levard C， Judy J D， et al. Fate of zinc oxide and silver nanoparticles in a pilot wastewater treatment plant and in processed biosolids[J]. Environ Sci Technol， 2013， 48 （1）： 104-112.
[32] Chandra V， Park J， Chun Y， et al. Water-dispersible magnetite-reduced graphene oxide composites for arsenic removal[J]. ACS Nano， 2010， 4（7）： 3979-3986.
[33] Lu W， Gao S， Wang J. One-pot synthesis of Ag/ZnO self-assembled 3D hollow microspheres with enhanced photocatalytic performance[J]. J Phys Chem C， 2008， 112（43）： 16792-16800.
[34] Wang Y， Zou B， Gao T， et al. Synthesis of orange-like Fe3O4/PPy composite microspheres and their excellent Cr （VI） ion removal properties[J]. J Mater Chem， 2012， 22（18）： 9034-9040.
[35] Kang D， Tong S， Yu X， et al. Template-free synthesis of 3D hierarchical amorphous aluminum oxide microspheres with broccoli-like structure and their application in fluoride removal[J]. RSC Adv， 2015， 5（25）： 19159-19165.
[36] Kang D， Yu X， Ge M， et al. One-step fabrication and characterization of hierarchical MgFe2O4 microspheres and their application for lead removal[J]. Micropor Mesopor Mater， 2015， 207： 170-178.
[37] Mohan D， Singh K P. Single-and multi-component adsorption of cadmium and zinc using activated carbon derived from bagasse—an agricultural waste[J]. Water Res， 2002， 36（9）： 2304-2318.
[38] Pan B， Xing B. Adsorption mechanisms of organic chemicals on carbon nanotubes[J]. Environ Sci Technol， 2008， 42（24）： 9005-9013.
[39] Yang X， Zhen M， Li G， et al. Preparation of Pd-decorated fullerenols on carbon nanotubes with excellent electrocatalytic properties in alkaline media[J]. J Mater Chem A， 2013， DOI：10.1039/c3ta11907f.
[40] Kang D， Yu X， Tong S， et al. Performance and mechanism of Mg/Fe layered double hydroxides for fluoride and arsenate removal from aqueous solution[J]. Chem Eng J， 2013： 228， 731-740. [41] Xiong R， Wang Y， Zhang X， et al. Facile synthesis of magnetic nanocomposites of cellulose@ ultrasmall iron oxide nanoparticles for water treatment[J]. RSC Adv， 2014， 4（43）： 22632-22641.
[42] Yu X， Kang D， Hu Y， et al. One-pot synthesis of porous magnetic cellulose beads for the removal of metal ions[J]. RSC Adv， 2014， 4（59）： 31362-31369.
[43] Tian Y， Wu M， Lin X， et al. Synthesis of magnetic wheat straw for arsenic adsorption[J]. J Hazard Mater， 2011， 193： 10-16.
[44] Liu R， Yu H， Huang Y. Structure and morphology of cellulose in wheat straw[J]. Cellulose， 2005， 12（1）： 25-34.
[45] Yu H， Liu R， Shen D， et al. Study on morphology and orientation of cellulose in the vascular bundle of wheat straw[J]. Polymer， 2005， 46（15）： 5689-5694.
[46] Cao M S， Yang J， Song W L， et al. Ferroferric oxide/multiwalled carbon nanotube vs polyaniline/ferroferric oxide/multiwalled carbon nanotube multiheterostructures for highly effective microwave absorption[J]. ACS Appl Mater Interfaces， 2012， 4（12）： 6949-6956.
[47] Tanada S， Kabayama M， Kawasaki N， et al. Removal of phosphate by aluminum oxide hydroxide[J]. J Coll Interf Sci， 2003， 257（1）： 135-140.
[48] Deng Y， Qi D， Deng C， et al. Superparamagnetic high-magnetization microspheres with an Fe3O4@SiO2 core and perpendicularly aligned mesoporous SiO2 shell for removal of microcystins[J]. J Am Chem Soc， 2008， 130（1）： 28-29.