Professional Documents
Culture Documents
www.elsevier.com/locate/jcis
Abstract
Monodisperse chitosan-bound Fe3 O4 nanoparticles were developed as a novel magnetic nano-adsorbent for the removal of heavy metal
ions. Chitosan was first carboxymethylated and then covalently bound on the surface of Fe3 O4 nanoparticles via carbodiimide activation.
Transmission electron microscopy micrographs showed that the chitosan-bound Fe3 O4 nanoparticles were monodisperse and had a mean
diameter of 13.5 nm. X-ray diffraction patterns indicated that the magnetic nanoparticles were pure Fe3 O4 with a spinel structure, and the
binding of chitosan did not result in a phase change. The binding of chitosan was also demonstrated by the measurement of zeta potential,
and the weight percentage of chitosan bound to Fe3 O4 nanoparticles was estimated to be about 4.92 wt%. The chitosan-bound Fe3 O4
nanoparticles were shown to be quite efficient for the removal of Cu(II) ions at pH > 2. In particular, the adsorption rate was so fast that
the equilibrium was achieved within 1 min due to the absence of internal diffusion resistance. The adsorption data obeyed the Langmuir
equation with a maximum adsorption capacity of 21.5 mg g−1 and a Langmuir adsorption equilibrium constant of 0.0165 L mg−1 . The pH
and temperature effects revealed that the adsorption capacity increased significantly with increasing pH at pH 2–5, and the adsorption process
was exothermic in nature with an enthalpy change of −6.14 kJ mol−1 at 300–330 K.
2004 Elsevier Inc. All rights reserved.
The preparations of chitosan-conjugated Fe3 O4 particles isopropanol) was added into the reaction mixture in drops
by a coprecipitation method, crosslinking method, and co- in 30 min. After reaction for 4 h at the same temperature,
valent binding method using coupling agents have been 200 ml of ethyl alcohol (70%) was added to stop the reac-
reported [31]. The resultant magnetic composites were ei- tion. Finally, the solid was filtered, rinsed with 70 and 99%
ther aggregated or unstable due to polymer crosslinking ethyl alcohol to desalt and dewater, and dried in an oven
or physisorption, the particle sizes were above 100 nm, at 50 ◦ C.
even larger than 1000 nm. So, following our previous work The binding of carboxymethyl chitosan was conducted
on PAA-bound magnetic nanoparticles, the preparation of following our previous work on the preparation of PAA-
monodisperse chitosan-bound Fe3 O4 nanoparticles was at- bound magnetic nanoparticles [17]. First, 100 mg of Fe3 O4
tempted in this study. Since chitosan has no suitable func- nanoparticles was added to 2 ml of buffer A (0.003 M phos-
tional groups to bind directly onto Fe3 O4 nanoparticles, it phate, pH 6, 0.1 M NaCl). Then, the reaction mixture was
was first carboxmethylated and then covalently bound onto sonicated for 10 min after adding 0.5 ml of carbodiimide
Fe3 O4 nanoparticles via carbodiimide activation. The sizes, solution (0.025 g ml−1 in buffer A). Finally, 2.5 ml of car-
structures, and surface charges of the product were charac- boxymethyl chitosan solution (50 mg ml−1 in buffer A) was
terized by transmission electron microscopy (TEM), X-ray added and the reaction mixture was sonicated for 60 min.
diffraction (XRD), and zeta potential meter. The capability The chitosan-bound Fe3 O4 nanoparticles were recovered
and behavior for the adsorption of heavy metal ions were in- from the reaction mixture by placing the bottle on a per-
vestigated using Cu(II) ions as a model compound. manent magnet with a surface magnetization of 6000 G.
The magnetic particles settled within 1–2 min and then were
washed with water and ethanol. NaCl was used for the
2. Materials and methods flocculation of magnetic nanoparticles. Its addition could
accelerate the magnetic separation, particularly in alkaline
2.1. Materials solutions [33].
Chitosan (low viscous), chloroacetic acid sodium salt, 2.3. Characterizations of chitosan-bound
o-phthaldialdehyde, β-mercaptoethanol, and ferrous chlo- Fe3 O4 nanoparticles
ride tetrahydrate were purchased from Fluka (Buchs). Car-
bodiimides (cyanamide, CH2 N2 ) were supplied from Sigma The amino groups on the chitosan-bound Fe3 O4 magnetic
Chemical Co. (St. Louis, MO). Ferric chloride, 6-hydrate nanoparticles were determined by spectrophotometric as-
was purchased from J.T. Baker (Phillipsburg). Glycine, say based on the reaction of o-phthaldialdehyde (OPA) and
natriumtetraborate-10-hydrate, and Cu(II) nitrate-2,5-hy- β-mercaptoethanol (BME) with primary amines [17,34].
drate were obtained from Riedel-deHaën (Seelze). Ammo- First, two reagents were freshly prepared as follows. For re-
nium hydroxide (29.6%) was supplied by TEDIA (Fairfield). action of primary amines of the solid, 300 µl ethanol solution
The water used throughout this work was the reagent-grade containing 0.25 M OPA and 300 µl of 0.1 M sodium borate
water produced by the Milli-Q SP ultrapure water purifica- (pH 9.5) containing 4% β-mercaptoethanol were added to
tion system of Nihon Millipore Ltd., Tokyo. All other chem- water to yield 50 ml of reagent 1 (R1). For the determina-
icals were the guaranteed or analytic grade reagents com- tion of excess OPA, 500 µl of 0.06 M glycine and 500 µl
mercially available and used without further purification. of 0.1 M sodium borate (pH 9.5) containing 0.5% BME
were added to water to yield 30 ml of reagent 2 (R2). Sec-
2.2. Preparation of chitosan-bound Fe3 O4 nanoparticles ondly, 20 mg of magnetic nanoparticles was incubated with
5 ml reagent R1 for 5 min at 25 ◦ C. The liquid solution
Fe3 O4 nanoparticles were prepared by coprecipitating was then separated from the magnetic nanoparticles via a
Fe2+ and Fe3+ ions by ammonia solution and treating un- permanent magnet. To 4 ml of 0.1 M sodium borate and
der hydrothermal conditions [17–20]. Ferric and ferrous 200 µl of reagent R2 in a test tube, 100 µl of supernatant
chlorides (molar ratio 2:1) were dissolved in water at a was added. After mixing by vortex for several minutes, the
concentration of 0.3 M iron ions. Chemical precipitation absorbance of the sample at 340 nm was determined on a
was achieved at 25 ◦ C under vigorous stirring by adding Hitachi U-3000 spectrophotometer. Finally, a reference so-
NH4 OH solution (29.6%). During the reaction process, the lution was prepared according to the above procedures but in
pH was maintained at about 10. The precipitates were heated the absence of magnetic nanoparticles. The concentration of
at 80 ◦ C for 30 min, washed several times with water and amino groups on the surface of magnetic nanoparticles was
ethanol, and then finally dried in a vacuum oven at 70 ◦ C. determined from the difference between the absorbances of
Chitosan was carboxymethylated according to the method reference solution and sample, using a molar absorption co-
of Chen and Park [32]. First, 3 g chitosan and 15 g sodium efficient ε = 6623 M−1 cm−1 obtained from the calibration
hydroxide were added into 100 ml of isopropanol/water curve.
(80/20) mixture at 60 ◦ C to swell and alkalize for 1 h. Then, The size and morphology of magnetic nanoparticles
20 ml of monochloroacetic acid solution (0.75 g ml−1 in were observed by TEM using a JEOL Model JEM-1200EX
448 Y.-C. Chang, D.-H. Chen / Journal of Colloid and Interface Science 283 (2005) 446–451
Fig. 3. Zeta potentials of the naked (P) and chitosan-bound (!) Fe3 O4 Fig. 5. Effect of pH on the adsorption of Cu(II) ions by magnetic nano-
nanoparticles at different pH. adsorbent. Initial Cu(II) ion concentration: 200 (!) and 1150 mg L−1 (P).
Fig. 4. Dependence of adsorption capacity on contact time. Fig. 6. Equilibrium isotherm for the adsorption of Cu(II) ions on magnetic
nano-adsorbent at pH 5 and 300 K. The inset illustrates the linear depen-
dence of Ce /q on Ce .
the literature value of 6.5 [35]. After chitosan bonding, the
pI was shifted to 5.95. This also confirms the binding of
pH > 5, white Cu(OH)2 precipitate formed and so no ad-
chitosan and reveals that the chitosan-bound Fe3 O4 nanopar-
sorption experiments were performed. Noteworthly, almost
ticles were positively charged at pH < 5.95.
no Cu(II) ions were adsorbed at pH 2. This condition might
be utilized for the desorption of Cu(II) ions from the mag-
3.4. Adsorption properties netic nano-adsorbent.
The equilibrium isotherm for the adsorption of Cu(II)
The feasibility of chitosan-bound Fe3 O4 nanoparticles ions on magnetic nano-adsorbent at pH 5 and 300 K is shown
as a magnetic nano-adsorbent for the removal of heavy in Fig 6. The adsorption data were analyzed according to the
metal ions from aqueous solutions was demonstrated using linear form of the Langmuir isotherm as follow [36],
Cu(II) ions as a model compound. First, the time required
Ce 1 Ce
to achieve adsorption equilibrium was determined. The typ- = + , (1)
ical experiments were conducted at pH 5, 300 K, and an q Kqm qm
initial Cu(II) ion concentration of 1100 mg L−1 . As shown where q is the adsorption capacity (mg g−1 ) based on the
in Fig. 4, the adsorption capacities (q) for the contact time dry weight of nano-adsorbent, Ce is the equilibrium concen-
of 1 to 60 min had no significant differences. This reveals tration (mg L−1 ) in solution, qm is the maximum adsorption
that the adsorption equilibrium was reached within 1 min. capacity (mg g−1 ), and K is the Langmuir adsorption equi-
Such a fast adsorption rate could be attributed to the absence librium constant (L mg−1 ). As shown in the inset in Fig. 6,
of internal diffusion resistance, like the PAA-based magnetic the plot of Ce /q vs Ce at various temperatures yielded
nano-adsorbent developed in our previous work. straight lines, revealing that the adsorption of Cu(II) ions
The effect of pH on the adsorption of Cu(II) ions by on magnetic nano-adsorbent obeys the Langmuir adsorption
the magnetic nano-adsorbent at 300 K and the initial Cu(II) isotherm. From the slope and intercept, the values of qm and
ion concentrations of 200 and 1150 mg L−1 is illustrated in K might be estimated as 21.5 mg g−1 and 0.0165 L mg−1 ,
Fig. 5. It was found that the adsorption capacity increased respectively. It was notable that only 4.92 wt% of chitosan
with increasing the solution pH at an initial Cu(II) ion con- was responsible for the adsorption of Cu(II) ion. The max-
centration of 1150 mg L−1 . This might be due to the less imum adsorption capacity based on the weight of chitosan
insignificant competitive adsorption of hydrogen ions. At a was 437 mg g−1 , which was significantly higher than the
lower initial Cu(II) ion concentration (200 mg L−1 ), the ad- value 284 mg g−1 observed for the chitosan-coated silica
sorption capacity increased with increasing the solution pH beads by Shi et al. [37] and those values below 200 mg g−1
and then remained unchanged at pH 3–5. This might be due usually observed for the chitosan beads [30,38]. This might
to the fact that almost all Cu(II) ions were adsorbed. At be reasonably referred to the high specific surface area be-
450 Y.-C. Chang, D.-H. Chen / Journal of Colloid and Interface Science 283 (2005) 446–451
[28] G. Karthikeyan, K. Anbalagan, N.M. Andal, J. Chem. Sci. 116 (2004) [34] V.G. Janolino, H.E. Swaisgood, Appl. Biochem. Biotechnol. 36 (1992)
119. 81.
[29] R.S. Juang, H.J. Shao, Water. Res. 36 (2002) 2999. [35] I. Iwasaki, S.R.B. Cooke, Y.S. Kim, SME Trans. 223 (1962) 113.
[30] J.C.Y. Ng, W.H. Cheung, G. McKay, J. Colloid Interface Sci. 255 [36] W.S.W. Ngah, C.S. Endud, R. Mayanar, Funct. Polym. 50 (2002)
(2002) 64. 181.
[31] H. Honda, A. Kawabe, M. Shinkai, T. Kobayashi, J. Ferment. Bio- [37] Q.H. Shi, Y. Tian, X.Y. Dong, S. Bai, Y. Sun, Biochem. Eng. J. 16
eng. 86 (1998) 191. (2003) 317.
[32] X.G. Chen, H.J. Park, Carbohydr. Polym. 53 (2003) 355. [38] R. Navarro, J. Guzmán, I. Saucedo, J. Revilla, E. Guibal, Macromol.
[33] A. Kondo, H. Fukuda, J. Ferment. Bioeng. 84 (1997) 337. Biosci. 3 (2003) 552.