Carboxymethyl-β-cyclodextrin Modified Magnetic Nanoparticles for Effective Removal of Arsenic from Drinking Water: Synthesis and Adsorption Studies

INTRODUCTION

Arsenic contamination, which is present mainly as oxyanion compounds in drinking water and groundwater, is a serious environmental worldwide problem because of its toxicity and health hazards. The ingestion of arsenic can result in both cancerous and non-cancerous health effects such as disturbance of cardiovascular and nervous system function, pigmentation, depigmentation, skin cancer and cancer of internal organs [1].

Arsenic exists in four oxidation states, +5 (arsenate), +3 (arsenite), 0 (arsenic), and -3 (arsine). The toxicity of arsenic depends on its speciation, For example, arsenite is significantly more toxic than arsenate. It is typically more difficult to remove arsenite than arsenate from contaminated water. Under normal pH conditions (in natural waters), arsenite is mostly found as an uncharged species (H₃AsO₃), and negatively charged species (H₂AsO₃^- , HAsO₃^2-and AsO₃^3-) are found only at high pH [2].

Arsenic contamination of water results from both natural and anthropogenic activities. Arsenic is present in water due to dissolution of minerals as well as human activities such as mining, smelting of metal ores and use of pesticides causing arsenic pollution [3]. Reports have indicated arsenic pollution in many regions of countries around the world including USA, China, Chile, Bangladesh, Taiwan, Mexico, Argentina, Poland, Canada, Japan, India and Islamic Republic of Iran [4,5]. Due to the negative impacts of arsenic on human health, the US Environmental Protection Agency, the World Health Organization (WHO), and the European Commission have reduced the maximum contaminant level of arsenic in drinking water from 50 to 10 ppb. However, some countries have kept the earlier WHO guideline of 50 ppb as their standard [3,6]. Many arsenic removal technologies such as precipitation [7], membrane separation [8,9], nanofiltration [10,11], coagulation [12], ion exchange [13], microbial transformation [14,15], and adsorption [16,17,18] have been developed. Most of these techniques are only effective in removing arsenate, and they require a pretreatment step for oxidation of arsenite to arsenate. These pretreatment chemical methods can be used easily; however, their applications are limited by producing large amount of toxic sludge, and they need further treatment before disposal into the environment [19].

Among the removal methods, adsorption from solution has received more attention due to its low cost and high efficiency for arsenic treatment [20]. Many adsorbent materials have been used including activated alumina, activated carbon, red mud, bauxsol, etc [21]. Therefore, one of the key factors in adsorption-based technologies is the development of highly effective and inexpensive adsorbents [22]. Recently, adsorption through Fe₃O₄ magnetic nano-sized particles has been more popular because of the specific characteristics of these materials. Particular properties of them such as extreme small size, high surface-area-to-volume ratio and the absence of internal diffusion resistance, provides better kinetics for adsorption of arsenic from aqueous solution. Magnetic nano-adsorbents have the qualities of both magnetic separation techniques and nano-sized materials, thus they can be easily separated with an external magnetic field [23, 24]. However, these adsorbents are not stable and hardly can be recycled because Fe₃O₄ is highly susceptible to oxidation when it is exposed to the atmosphere due to its small size [25]. In order to improve the stability and functionality, the iron oxide nanoparticles are often modified with natural or synthetic polymers.

Cyclodextrins are polysaccharides produced through the degradation of starch by cyclodextrin glucanotransferase enzyme. β-cyclodextrin (β-CD) is a cyclic oligosaccharide consist of seven α-D-glucose units connected through α-(1,4) linkages. The structure of these molecules is like toroidal truncated cones containing a lipophilic cavity with two hydroxyl groups one lying on the outside and the other lying in the inside [26,25]. Cyclodextrins are available on a large scale. Their production is not costly, and most importantly they have the ability to form inclusion complexes (guest-host complexes) with a wide variety of organic and inorganic compounds in its hydrophobic cavity. In addition, metal ions can be complexed by cyclodextrins through hydroxyl groups [27].

In this study, a magnetic nano-adsorbent was fabricated by surface modification of Fe₃O₄ nanoparticles with β-CD for the adsorption of arsenic contamination. The size, morphology and properties of the β-CD modified magnetic nanoparticles were characterized using different analytical tools. The mean diameter of β-CD modified Fe₃O₄ magnetic nanoparticles were ~10 nm. These modified nanoparticles exhibit superparamagnetic properties at room temperature; therefore, they can be easily separated by applied magnetic field. They were evaluated as absorbents to remove arsenic, and the effects of several factors such as, pH, initial arsenic concentration, contact time, and mechanism of arsenic adsorption onto β-CD modified magnetic nanoparticles were investigated in this report. In Iran, in western and north-western provinces especially in Kurdistan province arsenic contamination has been reported in groundwater sources [5,28]. Real water samples from Kurdistan villages were collected, and the ability of modified magnetic nanoparticles in removing arsenic in real samples was examined. Overall, the major end of the current research is ascertaining the capability of β-CD modified magnetic nanoparticles to remove trivalent arsenic from water in a one-step operation with no pretreatment. The maximum removal efficiency of arsenic (III) at optimal conditions is about 85%.

EXPERIMENTAL

Materials

Iron (II) chloride tetrahydrate, iron (III) chloride hexahydrate, sodium hydroxide (NaOH), ammonium hydroxide (25%), β-cyclodextrin (β -CD), monochloroacetic acid, Carbodiimides (cyanamide, CH₂N₂), arsenic trioxide (As₂O₃) and methanol (99%) were purchased from Merck. All chemicals were the guaranteed or analytic grade reagents commercially available and used without further purification. Arsenic solutions that were used in different experiments were prepared by diluting a stock solution (1000 ppm of arsenic). The stock solution was prepared by solving 50 mg of As₂O₃ salt in 100 ml deionized water. For preparing working solution from 1 up to 100 ppm of arsenic, 0.01 ml to 1 ml of the stock solution were diluted in 10 ml deionized water.

Apparatus

In order to investigate the size and morphology of magnetic nanoparticles, transmission electron microscopy (CM10, Philips) was used. β-CD’s grafting onto the surface of magnetic nanoparticles (MNPs) was monitored by Fourier transform infrared spectroscopy (RX1, Perkin-Elmer). The crystalline structure of Fe₃O₄ nanoparticles was characterized by X-ray diffraction technique (D8 Advance, Bruker). The magnetic properties of the Fe₃O₄ nanoparticles coated with β-cyclodextrin were investigated using a vibrating sample magnetometer (Megnatis Daghigh Kavir, Iran). The size and size distribution of nanoparticles were measured by particle size analyzer (L-550, HORIBA).

Preparation of Nacked Fe₃O₄ magnetic nanoparticles

Fe₃O₄ nanoparticles were prepared by chemical co-precipitation method. A complete precipitation of Fe₃O₄ was achieved under alkaline condition, while maintaining a molar ratio 1:2 of Fe²⁺ and Fe³⁺. In a typical synthesis to obtain 1 g Fe₃O₄ precipitate, 0.86 g of FeCl₂·4H₂O and 2.36 g of FeCl₃·6H₂O were dissolved in 40 ml of distilled water with vigorous stirring at a speed of 1000 rpm. 5 ml of NH₄OH (25%) was added after the solution was heated to 80 ^oC. The reaction was continued for 30 min at 80 ^oC under constant stirring to ensure the complete growth of the nanoparticle crystals. The resulting particles were then washed with Distilled water at least 5 times to remove any unreacted chemicals and dried [24].

Preparation of Carboxymethyl β-cyclodextrin

Carboxymethyl-β-cyclodextrin (CM-βCD) was prepared following the previous procedure [29]. CM-βCD was synthesized in the alkaline condition by reacting monochloroacetic acid with β-CD. A mixture of β-CD (2 g) and NaOH (1.86g) in water (7.4 ml) was treated with a 16.3% monochloroacetic acid solution (5.4 ml) at 50 ^oC for 5 h. The temperature of the reaction mixture was reduced to room temperature, and then the pH values were adjusted in the range of 6–7. The obtained neutral solution was added to methanol, and the produced white precipitate, carboxymethylated b -cyclodextrin, was filtered and dried in oven at 50 ^oC.

Preparation of CM-βCD Modified Magnetic Nanoparticles

β-cyclodextrin modified magnetic nanoparticles can be prepared using one-step or two-step methods. By one-step co-precipitation method, in which iron precursors (Fe²⁺ and Fe³⁺) and CM-βCD were mixed together, the binding of CM-βCD onto the Fe₃O₄ surface was conducted. Briefly, 0.57 g of FeCl₂·4H₂O, 1.57 g FeCl₃·6H₂O and 1 g CM-βCD were dissolved in 26.7 ml of distilled water with vigorous stirring at a speed of 1200 rpm. When the temperature of the reaction mixture reached 90 ^oC , 3.5 ml of NH₄OH (25%) was added in drop. The reaction was maintained at 90 ^oC under constant stirring for 1 h. The resulting nanoparticles were then washed with distilled water to remove any unreacted chemicals and then it was dried in oven at 70 ^oC. In this method, the carboxyl groups of CM-βCD directly reacted with the surface OH groups on the magnetite to form Fe-carboxylate [29].

Adsorption Experiments

The adsorption experiments of As(III) by CM-βCD modified magnetic nanoparticles (CM-βCD-MNPs) were investigated in aqueous solutions. For each treatment, 30 mg CM-βCD-MNPs was added into 5mL of arsenic solution. The solutions of As(III) were obtained by dissolving arsenic trioxide (As₂O₃) into distilled water. The influence of experimental factors such as pH value, contact time, adsorbent dosage and initial arsenic concentration on the removal efficiency of As(III) was investigated. When equilibrium was achieved, magnetic nanoparticles were separated magnetically from arsenic solution using a magnet. The concentration values of As(III) were measured using the inductively coupled plasma spectroscopy (ICP) method. For comparison, the adsorption of As(III) by the naked Fe₃O₄ nanoparticles (unmodified MNPs) was also investigated.

The adsorption isotherm experiments were performed using different initial As(III) concentrations. The adsorption capacity (q_e, mg/g) and removal efficiency (E) of the adsorbent were calculated using the following equations:      

                                               (1)

                                         (2)

Where C₀ and C_e are initial and equilibrium concentration (mg/L) of As(III) solution, respectively; V is the volume of the As(III) solution; and m is the weight of the β-CD modified magnetic nanoparticles.

RESULTS AND DISCUSSIONS

One of the most important steps in magnetic nanoparticle preparation is their characterization. It means the investigation of their size, shape and size distribution and also, their magnetic properties. For this purpose different methods were used and the results were shown following.

Characterizations of β-cyclodextrin modified Fe₃O₄ nanoparticles

Transmission Electron Microscopy (TEM)

Transmission electron microscopy (TEM) was utilized to investigate the size and morphology of magnetic nanoparticle samples. As shown in Fig. 1, the sample includes small nanoparticles that are relatively uniform in size and shape. The average size of the particles is about 10 nm. This reveals that the binding process did not significantly result in changes in the size of the particles but it prevents agglomeration of nanoparticles which leads to lower distribution of modified nanoparticles rather than unmodified nanoparticles.

Fig.1 TEM images of the (a) naked and (b) CM-bCD modified Fe₃O₄ nanoparticles.

Dynamic Light Scattering (DLS)

In Fig. 2 the size distribution graphs for the naked and β-CD modified Fe₃O₄ nanoparticles obtained by DLS method are presented. The average diameters of the naked and β-CD modified magnetic nanoparticle were equal to 13.7 and 18.5 nm, respectively. It can be seen for Fig. 2 that, they are low-dispersed nanoparticles. As it evidence, the nanoparticles size observed by DLS is larger than those obtained by TEM because, β-CD coating increases the hydrodynamic diameter of nanoparticles [25].

Fig. 2 DLS plots of the (a) naked and (b) b-CD modified Fe₃O₄ nanoparticles.

FTIR analysis

The CM-β-CD binding on the surface of magnetic nanoparticles was confirmed by FTIR spectroscopy. FTIR spectra of naked MNPs, CM-β-CD-MNPs, and pure CM-β-CD in the range of 350–4000 cm^-1 wavenumbers are shown in Fig. 3.

In the spectra of the Fe₃O₄ nanoparticles, the absorption peak at 585  is the characteristic of Fe-O-Fe bond in Fe₃O₄. The broad band around 3426  is due to –OH stretching vibrations. The spectrum of CM-β-CD shows the characteristic peaks at 948, 1031 and 1190. The peak at 948 is due to the R-1,4-bond skeleton vibration of β-CD, and the peaks at 1031 and 1190 corresponded to the asymmetric glycosidic (C–O–C) vibrations and coupled υ(C–C/C–O) stretch vibration [21]. All of these characteristic peaks in the spectrum of CM-β-CD (900–1200) can also be seen in the spectrum of CM-β-CD-MNPs with slight differences. The characteristic peak appeared at 1606  is due to bands of -COOM groups (M represents metal ions), indicates that the -COOH groups of CM-β-CD reacted with the surface of Fe₃O₄ particles [30]. These findings indicate that the binding of CM-β-CD on the surface of Fe₃O₄ nanoparticles were done successfully.

Fig. 3 FTIR spectra of (a) naked and (b) b-CD modified Fe₃O₄ nanoparticles and (C) pure CM-bCD.

UV-Vis spectra

The UV–visible absorption spectra of the naked and β-CD modified nanoparticles are illustrated in Fig. 4. As can be seen in the absorption spectra reported in Fig. 4, after β-CD coating on the MNPs the UV-Vis spectrum of these nanoparticles was changed. This phenomenon can be also used as another confirmation for binding of β-CD on the Fe₃O₄ nanoparticles’ surface.

Fig. 4 UV–visible spectra of the naked (Dot line) and b-CD modified (Solid line) Fe₃O₄ nanoparticles.

XRD analysis

Fig.5 shows the XRD pattern of the CM-β-CD modified Fe₃O₄ nanoparticles. There are six characteristic peaks for Fe₃O₄ at 2θ values of 30.3˚, 35.6˚, 43.25˚, 53.65˚, 57.15˚, and 62.85˚ marked by their indices (2 2 0), (3 1 1), (4 0 0), (4 2 2), (5 1 1), and (4 4 0), respectively. These observations confirm the presence of inverse cubic spinel structure of the resultant nanoparticles [29]. It is also clear that the coating did not lead any phase change of Fe₃O₄ nanoparticles. The average size of the CM-β-CD modified Fe₃O₄ can be calculated from Scherrer equation:

                                       (3)

Where d is particle diameter, λ is X-ray wavelength, β is the peak width of half-maximum, and θ is Bragg’s diffraction angle in degree. The obtained size of CM-β-CD modified MNPs using strongest peak (3 1 1) at 2θ=35.6˚ and λ=1.54 A˚ was about 8.8 nm, which is in consistent with the results of TEM.

Fig. 5 XRD pattern of the b-CD modified Fe₃O₄ nanoparticles synthesized.

Vibrating sample magnetometer (VSM) analysis

In order to examine the magnetic properties of prepared nanoparticles, they were analyzed by vibrating sample magnetometer (VSM) at room temperature. The magnetization hysteresis loops of naked and CM-β-CD-MNPs were shown in Fig. 6. The saturated magnetizations for naked MNPs and CM-β-CD-MNPs are 54 and 48 emu. This demonstrates that all samples are superparamagnetic which can responsive to an applied magnetic field and retain no permanent magnetization after removing the applied magnetic field. The saturated magnetization of the Fe₃O₄ nanoparticles decreased after coating with CM-β-CD (Fig. 6). This is mainly attributed to the existence of non-magnetic materials on the surface of nanoparticles. Similar results are reported by Badruddoza et. al. [29]. Materials with superparamagnetic properties can be easily separated from the solution with the aid of an external magnetic field within several minutes.

Fig. 6 Magnetic hysteresis curves of naked (solid line) and b-CD modified (dot line) Fe₃O₄ nanoparticles.

Adsorption of arsenic onto β-CD modified MNP

The adsorption process can be influenced by some experimental parameters. The effects of these parameters should be investigated and the optimal conditions should be used for real samples analysis.

Effects of initial pH on arsenic adsorption

One of the most important parameters in arsenic removal process is the pH value of the sample. Since the pH variation can influence on the ionic state of the surface functional groups of nanoparticles, it can be effective on the adsorption process. On the other hand, the arsenic chemistry in water also highly depends on pH values. Therefore, arsenic adsorption ability of CM-β-CD-MNPs was investigated at room temperature by varying pH values in the range of 8 to 11 (Fig. 7). The removal efficiency of arsenic improved by increasing pH values and then remains almost constant at pH 9 -11 because of the pH-dependency of As (III). However, pH 9 was selected as the optimal value since no significant improvement in arsenic removal was observed between 9 to 11 pH values.

Fig. 7 The effect of pH on the arsenic removal efficiency by the b-CD modified Fe₃O₄ nanoparticle The removal studies were carried out by applying 6mg/ml b-CD-modified Fe₃O₄ on the As(III) solution with 10 ppm concentration, during 10 min contact time at room temperature).

Effect of Magnetic Nanoparticles Dosage

The effect of adsorbent dosage on arsenite adsorption capacity and removal efficiency was investigated and the results were shown in Fig. 8. It is evident that the removal efficiencies of arsenite increase with increasing adsorbent dosage, while the amount of adsorption capacity (q_e) decreases. The increase in the removal efficiency is due to an increase in the adsorbent amount which provides more adsorbent surface for the solute to be adsorbed.

 Fig. 8 The effect of b-CD-modified Fe₃O₄ nanoparticle adsorbent dosage on arsenic efficiencies and adsorption capacity.

Effect of Initial Concentration

As mentioned above, CM-βCD-MNPs provided good adsorption capacity for As(III) at pH 9 and 6 mg/ml adsorbent amount by applying 30 minutes contact time. The adsorption isotherm of arsenic was obtained by changing the initial concentration of As(III) at values ranging from 1 to 100 ppm. The removal efficiency of As(III) decreases with an increase in initial As(III) concentration (Fig. 9). This behavior was due to the fact that the total available adsorption sites for a fixed amount of adsorbent are limited, which leads to a decrease in removal percentage corresponding to an increased initial adsorbate concentration.