- Email: [email protected]
carbon nanocomposites and their adsorption property for methyl orange from aqueous solution
Journal of Colloid and Interface Science 389 (2013) 10–15
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Synthesis of mesoporous magnetic Co-NPs/carbon nanocomposites and their adsorption property for methyl orange from aqueous solution Peng Zhang, Qiao An, Jia Guo, Chang-Chun Wang ⇑ State Key Laboratory of Molecular Engineering of Polymers, Department of Macromolecular Science, and Laboratory of Advanced Materials, Fudan University, Shanghai 200433, People’s Republic of China
a r t i c l e
i n f o
Article history: Received 28 August 2011 Accepted 9 August 2012 Available online 24 August 2012 Keywords: Magnetic responsibility High saturation magnetization Mesoporous carbon Nanocomposites Dye adsorption/desorption
a b s t r a c t Mesoporous magnetic Co-NPs(nanoparticles)/carbon nanocomposites were synthesized by carbonization of polyacrylonitrile (PAN) microspheres entrapped with cobalt salt for the first time. The structure and morphology of the porous magnetic nanocomposites were characterized by X-ray diffraction (XRD), transmission electron microscopy (TEM), vibrating sample magnetometry (VSM), and N2 adsorption– desorption technique. The nanocomposites possess very high saturation magnetization (Ms is up to 133 emu/g), near-zero remanence, and very low coercivity (Hc is down to 0.023 KOe). Meanwhile, the nanocomposites have mesoporous structure with average pore size of 4 nm and high specific surface area of 232 m2/g, which can be tuned by changing the carbonization conditions. Using methyl orange (MO) as model pollutant in water, the mesoporous magnetic nanocomposites showed good adsorption capacity of 380 mg/g, and the absorbed MO could be easily released in ethanol. The mesoporous nanocomposites were facile separated from solution under external magnetic force, and over 85% adsorption capacity for MO could be retained after five adsorption/desorption cycles. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Organic dyes are widely used in the textile, paper, plastic, and leather industry to color commercial products. However, the organic dyes in industrial wastewater may result in serious environmental problems because of their toxicity and poor biodegradability, which may risk humankind’s health [1–4]. Besides, these dyes are very difficult to be removed from wastewater by chemical or biological degradation methods, due to their stability against heat, light, and many chemical reagents. Accordingly, it is necessary to find an efficient way to separate and remediate these organic pollutants. Of all the technologies applied in the removal of organic pollutants, adsorption of the organic dyes on activated carbon is shown to be the most effective and frequently used method due to its high surface area and fast adsorption kinetics [5–9]. However, its small particle size and high regeneration temperature sometimes cause problems in industrial applications; it is quite effort-taking to separate activated carbon materials from the treated water due to their small particle sizes. Over the past decades, magnetic separation technology has received increasing attention for its easy separation under magnetic field that also brings lots of advantages over the conventional materials [10,11]. To date, a series of porous magnetic composite materials have been successfully synthesized and used in many ⇑ Corresponding author. Fax: +86 21 65640293. E-mail address: [email protected] (C.-C. Wang). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2012.08.022
applications [12–18]. For examples, Xiao et al. reported a facile route for synthesis of porous composites consisting of iron carbide, iron, and graphite carbon [19]. Dong et al. synthesized mesostructured c-Fe2O3/C composites with c-Fe2O3 nanoparticles embedded in the wall of ordered mesoporous carbon materials by a co-casting method [20]. Recently, Lu et al. demonstrated a new kind of magnetically separable hydrogenation catalyst by using Co-containing mesoporous carbon [21]. However, most of the porous magnetic composite materials have low saturation magnetization values and the preparation procedure is usually rather complex, which may limit their practical applications. Therefore, the nanocomposites with large surface area, high saturation magnetization, and superparamagnetism are highly desired to be used as industrial sorbents. Cobalt has high magnetic moment, but cobalt nanoparticles (Co-NPs) are easily oxidized and lose their high magnetic moment significantly. According to our previous study, cobalt nanoparticles have been embedded in the interior of the carbon microspheres, which protected the metal nanocrystals from oxidation, the CoNPs doped carbon microspheres possessed very high saturation magnetization [22]. In the present work, we demonstrated a facile and controllable route to synthesize mesoporous magnetic carbon nanocomposites via the simple carbonization process of PAN microspheres entrapping Co2+ as cobalt precursor. These magnetic carbon nanocomposites combined the advantages of high surface area, high magnetization, and low coercivity. When methyl orange
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
11
was used as a model pollutant in water, it could be effectively adsorbed by the nanocomposites, and the adsorbed MO could be quickly released in ethanol. The mesoporous carbon nanocomposites could be easily separated by magnetic field during the adsorption and desorption cycles, which is of great importance for water treatment.
CoC-600–60 in ethanol lasted for 12 h, followed by removal of the ethanol solution under the magnet field. After the release operation, these CoC-600–60 particles were added in the MO solution under stirring for another adsorption.
2. Experimental section
Transmission electron microscopy (TEM) images were obtained on a Hitachi H-600 transmission electron microscope operating at 250 kV. Thermogravimetric analysis (TGA) measurements were carried out on a Perkin–Elmer Pyris-1 series thermal analysis system under a flowing nitrogen or air atmosphere at a scan rate of 20 °C/min from 100 to 700 °C. Powder X-ray diffraction (XRD) patterns were collected on a X’Pert Pro (Panalytical, The Netherlands) diffraction meter with Cu KR radiation at k = 0.154 nm operating at 40 kV and 40 mA. Nitrogen adsorption desorption measurements were performed on an ASAP2020 (Micromeritics, USA) accelerated surface area analyzer at 77 K under continuous adsorption conditions. BET and BJH analyses were used to determine the surface area, pore size, and pore volume. Magnetic characterization was carried out on a model 6000 physical property measurement system (Quantum Design, USA) at 300 K.
2.1. Materials Acrylonitrile was purchased from Fluka and divinyl benzene (DVB) was purchased from Aldrich. Sodium chloride, potassium persulfate, sodium nitrite, N,N-dimethylformamide (DMF), and cobalt(II) chloride were supplied by Sinopharm Chemical Reagent Co., Ltd., and all the reagents are analytical reagent grade. All the chemicals were used as received without further purification. The water was purified through a Millipore system. 2.2. Preparation of Co(II)/PAN microspheres The Co(II)/PAN microspheres were prepared by entrapping Co2+ into polyacrylonitrile (PAN) template microspheres. Monodisperse PAN-template microspheres were synthesized via soap-free emulsion polymerization as described previously [22]. In a typical synthesis, a solution of acrylonitrile (7.00 g), divinyl benzene (1.05 g), NaCl (0.004 g), and deionized water (133 mL) was charged into a flask, which was equipped with a mechanical stirring bar, a thermometer, and a condenser. The solution was purged with nitrogen to remove oxygen for 30 min and then heated to 70 °C. The initiator solution of potassium persulfate (0.070 g K2S2O8 in 7 mL water) was charged into the flask, and this reaction system continued heating for 8 h. The obtained microspheres were washed with deionized water by centrifugation for three times and then dispersed in DMF with a solid content of 1 wt% for further use. 11.90 g of CoCl2 was added to 53 mL PAN microspheres dispersion mentioned above and heated to 60 °C with mechanical stirring. A certain amount of NaNO2 was added into the solution and stirring for 24 h. The products were precipitated by deionized water and filtered, then washed several times, and dried off. 2.3. Preparation of mesoporous magnetic nanocomposites The mesoporous magnetic nanocomposites were prepared by the pyrolysis of the Co(II)/PAN microspheres. The as-prepared Co(II)/PAN microspheres were carbonized in a tube furnace under nitrogen atmosphere at a certain temperature for 1 h with a certain heating rate. The mesoporous magnetic nanocomposites were obtained after cooling. The samples were named as CoC-T–A (T is the carbonization temperature and A is the heating rate, °C/min).
2.5. Characterization
3. Results and discussion 3.1. Characterization of mesoporous magnetic nanocomposites The schematic procedure for the preparation of mesoporous magnetic Co-NPs/carbon nanocomposites is shown in Scheme 1. First, monodisperse PAN microspheres used as template were prepared via soup-free emulsion polymerization. Second, Co(II)/PAN microspheres were obtained by swelling the as-obtained PAN microspheres in CoCl2 DMF solution, and then, cobalt ions were coordinated with the cyano groups of PAN microspheres with the help of NaNO2. These microspheres were then precipitated and dried off, followed by a carbonization process under nitrogen atmosphere at high temperature to yield mesoporous Co-NPs/Carbon nanocomposites. Cobalt nanoparticles were formed and simultaneously entrapped inside the carbon microspheres via the thermal decomposition of the Co(II)-loading PAN microspheres under the reduction atmosphere provided by the carbonization of PAN. Fig. 1 shows the TEM images of PAN microspheres, Co(II)-loading PAN microspheres, and pryolyzed Co(II)/PAN. From Fig. 1a, we can find that the perfect spherical shape and narrow size distribu-
2.4. Methyl orange (MO) adsorption tests A certain amount of dried mesoporous magnetic naocomposites and 30 mL of a methyl orange (MO) solution (2 mg/mL) were mixed together, followed by stirring at room temperature for 12 h. The magnetic porous carbons with absorbed MO were separated by magnet, and the concentration of the solution was analyzed by UV–vis adsorption at the wavelength of 464 nm. The pH values of the solution were all near neutral in our adsorption experiments. For the recycling experiment, the sample of CoC-600–60 was attracted on the sidewall of the bottle by using magnet after adsorption of MO. The pollutant solution was removed and ethanol was added to the bottle. The release time for the pollutants to from
Scheme 1. Illustration of the synthetic procedure of mesoporous Co-NPs/carbon nanocomposites.
12
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
Fig. 1. TEM images of (a) PAN microspheres, (b) Co(II)-loading PAN microspheres, and (c) Co-NPs/carbon nanocomposites.
tion of the PAN microspheres were obtained via soap-free emulsion polymerization and the sizes of these microspheres are about 210 nm. Air-stable Co/carbon nanocomposites were prepared by entrapping CoCl2 in the DMF swollen PAN microspheres, followed by annealing these microspheres at high temperatures under nitrogen. After the swelling and cobalt-entrapping processes, the obtained Co(II)/PAN microspheres well maintained their shape (Fig. 1b) and PAN microsphere were used as an ideal template for Co2+ coordination. After pyrolysis at 800 °C, the obtained CoNPs/carbon nanocomposites possess good shape and rough surface (Fig. 1c). Due to the low contrast between the cobalt nanoparticles and carbon matrix, the cobalt nanoparticles cannot be observed directly in the TEM image. The crystalline structures of the carbonized products at different heating temperatures were characterized by powder XRD. As shown in Fig. 2a, Bragg reflections at 2h angles of 42.7°, 51.9°, and 76.3° can be observed when the heating temperatures were above 600 °C. The reflection peaks can be identified according to the databases of JCPDS-International Centre for Diffraction Data (ICDD), the reflection patterns well fit the metallic cobalt of fcc phase (ICDD-data file 15-0806). It is probably that the cobalt precursors cannot be reduced to zero valent cobalt if the carbonization temperature is lower than 600 °C, and cobalt nanoparticles could be formed above 600 °C. Calculating with the full width at halfmaximum of the diffraction peak (1 1 1) by the Debye–Scherrer formula, we could obtain that the crystal sizes of the cobalt nanoparticles are about 29.1, 30.4, and 33.9 nm for the samples of CoC-600–20, CoC-800–20, and CoC-1000–20. Fig. 2b shows the strongest saturation magnetizations of the mesoporous magnetic
nanocomposites, and a small magnetic hysteresis loop can be observed. The saturation magnetization value is about 133 emu/g, and the coercivity of the sample is about 0.023 KOe from the enlarged magnetization curves shown in the inset of Fig. 2b. This indicates that the mesoporous magnetic nanocomposites have superparamagnetic properties and can be separated quickly in liquid media by magnetic force due to their high saturation magnetization. Cobalt nanocrystals were formed after carbonization of the Co(II)/PAN microspheres. It can be explained that the pyrolysis of PAN microspheres promotes the reduction of cobalt precursors to zero-valent cobalt at high temperature under a nitrogen atmosphere, and the pyrolyzed atmosphere supplied the reducing condition for the cobalt precursors to elemental cobalt [23,24]. Besides, the NO formed form the thermal decomposition of NaNO2 may also take part in the reduction process. In order to investigate the influence of feeding ratio of NaNO2 to Co2+ on the cobalt loading amount, we kept all reaction conditions fixed and changed the amount of NaNO2 in the reaction recipes and the results were listed in Table 1. It can be seen that the loading amount of cobalt increases with the increasing NaNO2/Co2+ feeding ratio. When CoCl2 was dissolved in DMF, some kind of coordination compound formed in DMF, which may hinder Co2+ to coordinate with cyano-group [25]. NaNO2 will help Co2+ to coordinate with cyano-group as an additional ligand. Therefore, the addition of NaNO2 not only provides a part of reducing agent, but also improves the Co2+ loading amount. Fig. 3 shows the nitrogen adsorption–desorption isotherms and corresponding pore size distributions of the products carbonized at
Fig. 2. (a) XRD curves of the Co(II)-loading PAN microspheres pyrolyzed at different temperatures; (b) Magnetization curves of Co/C-1000–20 at room-temperature. The inset shows the enlarged magnetization curves.
13
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15 Table 1 The influence of feeding ratio of NaNO2 to Co2+ on the loading amount of Co(II) in PAN microspheres and carbon microspheres a.
a b
Sample code
Co/PAN-1
Co/PAN-2
Co/PAN-3
Co/PAN-4
Co2+/NaNO2 feeding ratio Co wt% in Co(II)/PANb Co wt% in Co-NPs/carbonb
1:0 3.7 4.9
1:1 4.3 6.3
1:3 28.6 42.1
1:6 48.2 71.0
In preparation, the molar feed ration of Co2+ to CN is 5:1. Calculated by TGA.
different temperatures, the data of the special surface area and pore volume are listed in Table 2. The resulting isotherm curves can be classified as a type IV isotherm, indicating the presence of mesoporous structure [26]. According to the de Boer’s classification, the hysteresis loop at P/P0 between 0.45 and 1.0 in Fig. 3a is characteristic of H4-type adsorption. This phenomenon may be attributed to the fact that the carbonized samples contain interconnected macropores and micropores [27]. The corresponding pore size distribution data calculated from the desorption batch of nitrogen isotherms by the BJH method show the similar narrow peak around 4 nm, which belongs to the mesopores. In addition, the hysteresis loop shifts to a higher relative pressure on approaching P/P0 = 1, indicating the presence of macropores. Both microporosity and macroporosity in the carbonized products may be formed via the removal of gaseous species, such as H2O, CO2 and NO, which are formed through the carbonization of PAN and decomposition of NaNO2 [19,28]. For comparison, two samples were prepared and characterized. If only PAN microspheres were carbonized at 800 °C under a nitrogen atmosphere (Table 2, sample C-800), the obtained product could be nonporous. However, the sample carbonized at 800 °C from pure PAN microspheres treated with NaNO2 shows the similar mesoporous properties (Table 2, sample C–NO2). Hence, it is highly probable that both the carbonization and the thermal decomposition process of NaNO2 play an important role for the formation of porosity. All samples showed the similar nitrogen adsorption–desorption isotherm curves, no matter at what temperature the Co(II)/PAN microspheres were carbonized. The surface areas decreased with the increasing carbonization temperature, which may be resulted form the increased shrinkage level with increasing the carbonization temperature. On the other hand, when the samples were carbonized under different heating rates at the same temperature, the corresponding results of the porous structure are listed in Table 3. It can be seen that the surface areas increase with the rising of heating rates. Besides, the volumes of micropores in samples increased at the
Table 2 Mesoporous properties of the nanocomposites carbonized at different temperatures. Sample code
CoC1000–20
CoC800–20
CoC600–20
CoC400–20
C-800a
C-NO2b
SBET (m2/g) Vp (cm3/g)
59.4 0.139
102.1 0.143
166.7 0.251
155.4 0.175
0.6 0.012
135.2 0.114
a
Sample obtained by pyrolysis of the pure PAN microspheres. Sample obtained by pyrolysis of the product prepared under the same condition of CoC-800–20 microspheres, but without CoCl2. b
Table 3 Textural parameters of the nanocomposites carbonized under different heating rates.
SBET (m2/g) Vp (cm3/g) Vmicro (cm3/g)a a
CoC600–20
CoC600–40
CoC600–60
CoC800–20
CoC800–40
CoC800–60
166.7 0.251 0.042
180.4 0.341 0.060
200.1 0.291 0.078
102.1 0.143 0.014
167.7 0.135 0.063
232.5 0.269 0.077
Micropore volume of the nanocomposites calculated by t-plot method.
same time. It is probably because that under the rising heating rates, the removal of gaseous species was more intensely to form the narrower size pores, which also enlarged the surface areas. These results indicate that the pore structure can be tuned by changing the carbonization temperature and heating rates. Sample carbonized at 600 °C with a heating rate of 60 °C/min (CoC-600– 60) shows a specific surface area of 201 m2/g, which was chose as the absorbent in the methyl orange adsorption test.
3.2. Methyl orange adsorption on mesoporous magnetic nanocomposites The treatment of dye in wastewater is very important for the protection of environment [29]. With the large surface area and excellent magnetic properties, the mesoporous magnetic nanocomposites should become an excellent adsorbent for removal of pollutants in water by adsorption. As a model pollutant, MO was chosen as an adsorbate for the adsorption–desorption experiments. As shown in Fig. 4, the color of the solution changed from orange to colorless after addition of the mesoporous magnetic nanocomposites to the MO solution. The nanocomposites with adsorbed pollutant were quickly separated by magnet and the solution became colorless. By stirring in ethanol, the absorbed MO could be released from the nanocomposites easily.
Fig. 3. (a) Nitrogen adsorption–desorption isotherms and (b) corresponding pore size distributions of mesoporous magnetic nanocomposites carbonized at different temperatures, the heating rate was 20 °C/min.
14
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
Fig. 4. Photographs for removal of methyl orange (MO) in water by mesoporous magnetic nanocomposites CoC-600–60: (a) MO water solution; (b) CoC-600–60 was added into the solution to adsorb the MO; (c) CoC-600–60 was separated from water by magnetic field and the solution became colorless; (d) MO was released after CoC-600–60 with adsorbed MO was transferred into ethanol.
Fig. 5. (a) Equilibrium adsorption isotherms and (b) langmuir plot for MO adsorption by CoC-600–60.
Fig. 6. MO adsorption amounts of CoC-600–60 and CoC-1000–20 as a function of time.
Fig. 7. Relative adsorption capacity of CoC-600–60 for MO in water at room temperature for different cycles.
To evaluate the adsorption capacities of the as-obtained mesoporous magnetic nanocomposites for MO, the adsorption isotherm was conducted, as shown in Fig. 5. Various adsorption models have been proposed so far to study many kinds of adsorption systems. The adsorption behavior in our system has been successfully
described by Langmuir model. For this, the following Langmuir equation was employed [30]:
Qe ¼
Q mCe ðK d þ C e Þ
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
where Qm (mg/g) is the maximum adsorption capacity, and Kd is the dissociation constant. The plots of Ce/Qe against the equilibrium concentration Ce are shown in Fig. 5b. A good linear relationship is obtained for MO adsorption (R2 = 0.9962), which also confirms that the sorption followed the Langmuir model. As 1/Qm is the slope of the straight line in Fig. 5b, it can be estimated that the maximum MO adsorption capacity of CoC-600–60 is about 380 mg/g. Fig. 6 illustrates the MO adsorption amounts of CoC-600–60 and CoC-1000–20 as a function of time. It can be seen that the adsorption amount of MO increases with the time elongation and it is very quickly at the beginning and reached equilibrium for about 2 h, which demonstrates the fast adsorption rate of the magnetic porous nanocaomposites. Interestingly, the absorbed MO can be released almost completely from the mesoporous magnetic nanocomposites in ethanol. Due to the good regeneration ability and excellent magnetic separation properties, it is possible to quickly remove pollutants in water by adsorption–desorption cycle. Fig. 7 shows the recycling of CoC-600–60 nanocompaoites for the adsorption of MO in water. After five cycles, the mesoporous magnetic nanocomposites still retained the adsorption level over 85% for the pollutant.
4. Conclusions In summary, we have prepared magnetically separable mesoporous nanocomposites via the carbonization of PAN microspheres entrapped with cobalt precursors. During the carbonization process, the cobalt nanocrystals were formed and then protect by the in situ formed carbon from oxidation. The Co-NPs/ carbon nanocomposite exhibits high magnetization and low coercivity, and they can be easily separated from the solution by magnetic field. Besides, the Co-NPs/carbon nanocomposites have high surface area and a mesoporous structure, which can be tuned by controlling the carbonization temperature and heating rates. The synthesized mesoporous magnetic nanocomposites showed an excellent adsorbed ability to MO (380 mg/g), and the adsorbed MO can be easily released in ethanol. Therefore, it is expected to be a promising adsorbent for dyes in industrial wastewater.
15
Acknowledgments This work was supported by National Science Foundation of China (Grant nos. 20974023, 21034003, 51073040 and 21128001). References [1] E.J. Weber, R.L. Adams, Environ. Sci. Technol. 29 (1995) 1163–1170. [2] B. Royer, N.F. Cardoso, E.C. Lima, V.S.O. Ruiz, T.R. Macedo, C. Airoldi, J. Colloid Interf. Sci. 336 (2009) 398–405. [3] Z. Aksu, J. Yener, Waste Manage. 21 (2001) 695–702. [4] F. Herrera, A. Lopez, G. Mascolo, P. Albers, J. Kiwi, Appl. Catal. B: Environ. 29 (2001) 147–162. [5] A.W.M. Ip, J.P. Barford, G. McKay, J. Colloid Interf. Sci. 337 (2009) 32–38. [6] T. Otowa, Y. Nojima, T. Miyazaki, Carbon 35 (1997) 1315–1319. [7] R.-L. Tseng, F.-C. Wu, R.-S. Juang, Carbon 41 (2003) 487–495. [8] K. Mohanty, M. Jha, B.C. Meikap, M.N. Biswas, Ind. Eng. Chem. Res. 44 (2005) 4129–4138. [9] B. Özkaya, J. Hazard. Mater. 129 (2006) 158–163. [10] H.M. Chen, P.K. Chu, J.H. He, T. Hu, M.Q. Yang, J. Colloid Interf. Sci. 359 (2011) 68–74. [11] R. Afkhami, J. Hazard. Mater. 174 (2010) 398–403. [12] M. Fröba, R. Köhn, G. Bouffaud, Chem. Mater. 11 (1999) 2858–2865. [13] A.B. Fuertes, P. Tartaj, Chem. Mater. 18 (2006) 1675–1679. [14] D.W. Wang, F. Li, G.Q. Lu, H.-M. Cheng, Carbon 46 (2008) 1593–1599. [15] M. Sevilla, T. Valdés-Solís, P. Tartaj, A.B. Fuertes, J. Colloid Interf. Sci. 340 (2009) 230–236. [16] Y.F. Zhu, E. Kockrick, S. Kaskel, T. Ikoma, N. Hanagata, J. Phys. Chem. C 113 (2009) 5998–6002. [17] S.M. Zhu, D. Zhang, Z.X. Chen, Y.M. Zhang, J. Mater. Chem. 19 (2009) 7710– 7715. [18] J. Jang, H. Yoon, Small 1 (2005) 1195–1199. [19] Z.H. Sun, L.F. Wang, P.P. Liu, S.C. Wang, B. Sun, D.Z. Jiang, F.S. Xiao, Adv. Mater. 18 (2006) 1968–1971. [20] X.P. Dong, H.R. Chen, W.R. Zhao, X. Li, J.L. Shi, Chem. Mater. 19 (2007) 3484– 3490. [21] A.H. Lu, W. Schmidt, N. Matoussevitch, H. Bönnemann, B. Spliethoff, B. Tesche, E. Bill, W. Kiefer, F. Schüth, Nanoengineering of a magnetically separable hydrogenation catalyst, Angew. Chem., Int. Ed. 43 (2004) 4303–4306. [22] Y.W. Chu, P. Zhang, J.H. Hu, W.L. Yang, C.C. Wang, J. Phys. Chem. C 113 (2009) 4047–4052. [23] J. Jang, H. Yoon, Adv. Mater. 15 (2003) 2088–2091. [24] T. Yogo, S. Naka, J. Mater. Sci. 24 (1989) 2115–2119. [25] C.Q. Cui, S.P. Jiang, A.C.C. Tseung, J. Electrochem. Soc. 138 (1991) 94–100. [26] Y.Q. Wang, C.M. Yang, W. Schmidt, B. Spliethoff, E. Bill, F. Schüth, Adv. Mater. 17 (2005) 53–56. [27] V. Mayagoitia, F. Rojas, I. Kornhauser, Pore network interactions in ascending processes relative to capillary condensation, J. Chem. Soc., Faraday Trans. 81 (1985) 2931–2940. [28] J.S. Hummelshøj, R.Z. Sørensen, M.Y. Kustova, T. Johannessen, J.K. Nørskov, C.H. Christensen, J. Am. Chem. Soc. 128 (2006) 16–17. [29] S. Shin, H. Yoon, J. Jang, Catal. Commun. 10 (2008) 178–182. [30] I. Langmuir, J. Am. Chem. Soc. 40 (1918) 1361–1403.
Contents lists available at SciVerse ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Synthesis of mesoporous magnetic Co-NPs/carbon nanocomposites and their adsorption property for methyl orange from aqueous solution Peng Zhang, Qiao An, Jia Guo, Chang-Chun Wang ⇑ State Key Laboratory of Molecular Engineering of Polymers, Department of Macromolecular Science, and Laboratory of Advanced Materials, Fudan University, Shanghai 200433, People’s Republic of China
a r t i c l e
i n f o
Article history: Received 28 August 2011 Accepted 9 August 2012 Available online 24 August 2012 Keywords: Magnetic responsibility High saturation magnetization Mesoporous carbon Nanocomposites Dye adsorption/desorption
a b s t r a c t Mesoporous magnetic Co-NPs(nanoparticles)/carbon nanocomposites were synthesized by carbonization of polyacrylonitrile (PAN) microspheres entrapped with cobalt salt for the first time. The structure and morphology of the porous magnetic nanocomposites were characterized by X-ray diffraction (XRD), transmission electron microscopy (TEM), vibrating sample magnetometry (VSM), and N2 adsorption– desorption technique. The nanocomposites possess very high saturation magnetization (Ms is up to 133 emu/g), near-zero remanence, and very low coercivity (Hc is down to 0.023 KOe). Meanwhile, the nanocomposites have mesoporous structure with average pore size of 4 nm and high specific surface area of 232 m2/g, which can be tuned by changing the carbonization conditions. Using methyl orange (MO) as model pollutant in water, the mesoporous magnetic nanocomposites showed good adsorption capacity of 380 mg/g, and the absorbed MO could be easily released in ethanol. The mesoporous nanocomposites were facile separated from solution under external magnetic force, and over 85% adsorption capacity for MO could be retained after five adsorption/desorption cycles. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Organic dyes are widely used in the textile, paper, plastic, and leather industry to color commercial products. However, the organic dyes in industrial wastewater may result in serious environmental problems because of their toxicity and poor biodegradability, which may risk humankind’s health [1–4]. Besides, these dyes are very difficult to be removed from wastewater by chemical or biological degradation methods, due to their stability against heat, light, and many chemical reagents. Accordingly, it is necessary to find an efficient way to separate and remediate these organic pollutants. Of all the technologies applied in the removal of organic pollutants, adsorption of the organic dyes on activated carbon is shown to be the most effective and frequently used method due to its high surface area and fast adsorption kinetics [5–9]. However, its small particle size and high regeneration temperature sometimes cause problems in industrial applications; it is quite effort-taking to separate activated carbon materials from the treated water due to their small particle sizes. Over the past decades, magnetic separation technology has received increasing attention for its easy separation under magnetic field that also brings lots of advantages over the conventional materials [10,11]. To date, a series of porous magnetic composite materials have been successfully synthesized and used in many ⇑ Corresponding author. Fax: +86 21 65640293. E-mail address: [email protected] (C.-C. Wang). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2012.08.022
applications [12–18]. For examples, Xiao et al. reported a facile route for synthesis of porous composites consisting of iron carbide, iron, and graphite carbon [19]. Dong et al. synthesized mesostructured c-Fe2O3/C composites with c-Fe2O3 nanoparticles embedded in the wall of ordered mesoporous carbon materials by a co-casting method [20]. Recently, Lu et al. demonstrated a new kind of magnetically separable hydrogenation catalyst by using Co-containing mesoporous carbon [21]. However, most of the porous magnetic composite materials have low saturation magnetization values and the preparation procedure is usually rather complex, which may limit their practical applications. Therefore, the nanocomposites with large surface area, high saturation magnetization, and superparamagnetism are highly desired to be used as industrial sorbents. Cobalt has high magnetic moment, but cobalt nanoparticles (Co-NPs) are easily oxidized and lose their high magnetic moment significantly. According to our previous study, cobalt nanoparticles have been embedded in the interior of the carbon microspheres, which protected the metal nanocrystals from oxidation, the CoNPs doped carbon microspheres possessed very high saturation magnetization [22]. In the present work, we demonstrated a facile and controllable route to synthesize mesoporous magnetic carbon nanocomposites via the simple carbonization process of PAN microspheres entrapping Co2+ as cobalt precursor. These magnetic carbon nanocomposites combined the advantages of high surface area, high magnetization, and low coercivity. When methyl orange
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
11
was used as a model pollutant in water, it could be effectively adsorbed by the nanocomposites, and the adsorbed MO could be quickly released in ethanol. The mesoporous carbon nanocomposites could be easily separated by magnetic field during the adsorption and desorption cycles, which is of great importance for water treatment.
CoC-600–60 in ethanol lasted for 12 h, followed by removal of the ethanol solution under the magnet field. After the release operation, these CoC-600–60 particles were added in the MO solution under stirring for another adsorption.
2. Experimental section
Transmission electron microscopy (TEM) images were obtained on a Hitachi H-600 transmission electron microscope operating at 250 kV. Thermogravimetric analysis (TGA) measurements were carried out on a Perkin–Elmer Pyris-1 series thermal analysis system under a flowing nitrogen or air atmosphere at a scan rate of 20 °C/min from 100 to 700 °C. Powder X-ray diffraction (XRD) patterns were collected on a X’Pert Pro (Panalytical, The Netherlands) diffraction meter with Cu KR radiation at k = 0.154 nm operating at 40 kV and 40 mA. Nitrogen adsorption desorption measurements were performed on an ASAP2020 (Micromeritics, USA) accelerated surface area analyzer at 77 K under continuous adsorption conditions. BET and BJH analyses were used to determine the surface area, pore size, and pore volume. Magnetic characterization was carried out on a model 6000 physical property measurement system (Quantum Design, USA) at 300 K.
2.1. Materials Acrylonitrile was purchased from Fluka and divinyl benzene (DVB) was purchased from Aldrich. Sodium chloride, potassium persulfate, sodium nitrite, N,N-dimethylformamide (DMF), and cobalt(II) chloride were supplied by Sinopharm Chemical Reagent Co., Ltd., and all the reagents are analytical reagent grade. All the chemicals were used as received without further purification. The water was purified through a Millipore system. 2.2. Preparation of Co(II)/PAN microspheres The Co(II)/PAN microspheres were prepared by entrapping Co2+ into polyacrylonitrile (PAN) template microspheres. Monodisperse PAN-template microspheres were synthesized via soap-free emulsion polymerization as described previously [22]. In a typical synthesis, a solution of acrylonitrile (7.00 g), divinyl benzene (1.05 g), NaCl (0.004 g), and deionized water (133 mL) was charged into a flask, which was equipped with a mechanical stirring bar, a thermometer, and a condenser. The solution was purged with nitrogen to remove oxygen for 30 min and then heated to 70 °C. The initiator solution of potassium persulfate (0.070 g K2S2O8 in 7 mL water) was charged into the flask, and this reaction system continued heating for 8 h. The obtained microspheres were washed with deionized water by centrifugation for three times and then dispersed in DMF with a solid content of 1 wt% for further use. 11.90 g of CoCl2 was added to 53 mL PAN microspheres dispersion mentioned above and heated to 60 °C with mechanical stirring. A certain amount of NaNO2 was added into the solution and stirring for 24 h. The products were precipitated by deionized water and filtered, then washed several times, and dried off. 2.3. Preparation of mesoporous magnetic nanocomposites The mesoporous magnetic nanocomposites were prepared by the pyrolysis of the Co(II)/PAN microspheres. The as-prepared Co(II)/PAN microspheres were carbonized in a tube furnace under nitrogen atmosphere at a certain temperature for 1 h with a certain heating rate. The mesoporous magnetic nanocomposites were obtained after cooling. The samples were named as CoC-T–A (T is the carbonization temperature and A is the heating rate, °C/min).
2.5. Characterization
3. Results and discussion 3.1. Characterization of mesoporous magnetic nanocomposites The schematic procedure for the preparation of mesoporous magnetic Co-NPs/carbon nanocomposites is shown in Scheme 1. First, monodisperse PAN microspheres used as template were prepared via soup-free emulsion polymerization. Second, Co(II)/PAN microspheres were obtained by swelling the as-obtained PAN microspheres in CoCl2 DMF solution, and then, cobalt ions were coordinated with the cyano groups of PAN microspheres with the help of NaNO2. These microspheres were then precipitated and dried off, followed by a carbonization process under nitrogen atmosphere at high temperature to yield mesoporous Co-NPs/Carbon nanocomposites. Cobalt nanoparticles were formed and simultaneously entrapped inside the carbon microspheres via the thermal decomposition of the Co(II)-loading PAN microspheres under the reduction atmosphere provided by the carbonization of PAN. Fig. 1 shows the TEM images of PAN microspheres, Co(II)-loading PAN microspheres, and pryolyzed Co(II)/PAN. From Fig. 1a, we can find that the perfect spherical shape and narrow size distribu-
2.4. Methyl orange (MO) adsorption tests A certain amount of dried mesoporous magnetic naocomposites and 30 mL of a methyl orange (MO) solution (2 mg/mL) were mixed together, followed by stirring at room temperature for 12 h. The magnetic porous carbons with absorbed MO were separated by magnet, and the concentration of the solution was analyzed by UV–vis adsorption at the wavelength of 464 nm. The pH values of the solution were all near neutral in our adsorption experiments. For the recycling experiment, the sample of CoC-600–60 was attracted on the sidewall of the bottle by using magnet after adsorption of MO. The pollutant solution was removed and ethanol was added to the bottle. The release time for the pollutants to from
Scheme 1. Illustration of the synthetic procedure of mesoporous Co-NPs/carbon nanocomposites.
12
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
Fig. 1. TEM images of (a) PAN microspheres, (b) Co(II)-loading PAN microspheres, and (c) Co-NPs/carbon nanocomposites.
tion of the PAN microspheres were obtained via soap-free emulsion polymerization and the sizes of these microspheres are about 210 nm. Air-stable Co/carbon nanocomposites were prepared by entrapping CoCl2 in the DMF swollen PAN microspheres, followed by annealing these microspheres at high temperatures under nitrogen. After the swelling and cobalt-entrapping processes, the obtained Co(II)/PAN microspheres well maintained their shape (Fig. 1b) and PAN microsphere were used as an ideal template for Co2+ coordination. After pyrolysis at 800 °C, the obtained CoNPs/carbon nanocomposites possess good shape and rough surface (Fig. 1c). Due to the low contrast between the cobalt nanoparticles and carbon matrix, the cobalt nanoparticles cannot be observed directly in the TEM image. The crystalline structures of the carbonized products at different heating temperatures were characterized by powder XRD. As shown in Fig. 2a, Bragg reflections at 2h angles of 42.7°, 51.9°, and 76.3° can be observed when the heating temperatures were above 600 °C. The reflection peaks can be identified according to the databases of JCPDS-International Centre for Diffraction Data (ICDD), the reflection patterns well fit the metallic cobalt of fcc phase (ICDD-data file 15-0806). It is probably that the cobalt precursors cannot be reduced to zero valent cobalt if the carbonization temperature is lower than 600 °C, and cobalt nanoparticles could be formed above 600 °C. Calculating with the full width at halfmaximum of the diffraction peak (1 1 1) by the Debye–Scherrer formula, we could obtain that the crystal sizes of the cobalt nanoparticles are about 29.1, 30.4, and 33.9 nm for the samples of CoC-600–20, CoC-800–20, and CoC-1000–20. Fig. 2b shows the strongest saturation magnetizations of the mesoporous magnetic
nanocomposites, and a small magnetic hysteresis loop can be observed. The saturation magnetization value is about 133 emu/g, and the coercivity of the sample is about 0.023 KOe from the enlarged magnetization curves shown in the inset of Fig. 2b. This indicates that the mesoporous magnetic nanocomposites have superparamagnetic properties and can be separated quickly in liquid media by magnetic force due to their high saturation magnetization. Cobalt nanocrystals were formed after carbonization of the Co(II)/PAN microspheres. It can be explained that the pyrolysis of PAN microspheres promotes the reduction of cobalt precursors to zero-valent cobalt at high temperature under a nitrogen atmosphere, and the pyrolyzed atmosphere supplied the reducing condition for the cobalt precursors to elemental cobalt [23,24]. Besides, the NO formed form the thermal decomposition of NaNO2 may also take part in the reduction process. In order to investigate the influence of feeding ratio of NaNO2 to Co2+ on the cobalt loading amount, we kept all reaction conditions fixed and changed the amount of NaNO2 in the reaction recipes and the results were listed in Table 1. It can be seen that the loading amount of cobalt increases with the increasing NaNO2/Co2+ feeding ratio. When CoCl2 was dissolved in DMF, some kind of coordination compound formed in DMF, which may hinder Co2+ to coordinate with cyano-group [25]. NaNO2 will help Co2+ to coordinate with cyano-group as an additional ligand. Therefore, the addition of NaNO2 not only provides a part of reducing agent, but also improves the Co2+ loading amount. Fig. 3 shows the nitrogen adsorption–desorption isotherms and corresponding pore size distributions of the products carbonized at
Fig. 2. (a) XRD curves of the Co(II)-loading PAN microspheres pyrolyzed at different temperatures; (b) Magnetization curves of Co/C-1000–20 at room-temperature. The inset shows the enlarged magnetization curves.
13
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15 Table 1 The influence of feeding ratio of NaNO2 to Co2+ on the loading amount of Co(II) in PAN microspheres and carbon microspheres a.
a b
Sample code
Co/PAN-1
Co/PAN-2
Co/PAN-3
Co/PAN-4
Co2+/NaNO2 feeding ratio Co wt% in Co(II)/PANb Co wt% in Co-NPs/carbonb
1:0 3.7 4.9
1:1 4.3 6.3
1:3 28.6 42.1
1:6 48.2 71.0
In preparation, the molar feed ration of Co2+ to CN is 5:1. Calculated by TGA.
different temperatures, the data of the special surface area and pore volume are listed in Table 2. The resulting isotherm curves can be classified as a type IV isotherm, indicating the presence of mesoporous structure [26]. According to the de Boer’s classification, the hysteresis loop at P/P0 between 0.45 and 1.0 in Fig. 3a is characteristic of H4-type adsorption. This phenomenon may be attributed to the fact that the carbonized samples contain interconnected macropores and micropores [27]. The corresponding pore size distribution data calculated from the desorption batch of nitrogen isotherms by the BJH method show the similar narrow peak around 4 nm, which belongs to the mesopores. In addition, the hysteresis loop shifts to a higher relative pressure on approaching P/P0 = 1, indicating the presence of macropores. Both microporosity and macroporosity in the carbonized products may be formed via the removal of gaseous species, such as H2O, CO2 and NO, which are formed through the carbonization of PAN and decomposition of NaNO2 [19,28]. For comparison, two samples were prepared and characterized. If only PAN microspheres were carbonized at 800 °C under a nitrogen atmosphere (Table 2, sample C-800), the obtained product could be nonporous. However, the sample carbonized at 800 °C from pure PAN microspheres treated with NaNO2 shows the similar mesoporous properties (Table 2, sample C–NO2). Hence, it is highly probable that both the carbonization and the thermal decomposition process of NaNO2 play an important role for the formation of porosity. All samples showed the similar nitrogen adsorption–desorption isotherm curves, no matter at what temperature the Co(II)/PAN microspheres were carbonized. The surface areas decreased with the increasing carbonization temperature, which may be resulted form the increased shrinkage level with increasing the carbonization temperature. On the other hand, when the samples were carbonized under different heating rates at the same temperature, the corresponding results of the porous structure are listed in Table 3. It can be seen that the surface areas increase with the rising of heating rates. Besides, the volumes of micropores in samples increased at the
Table 2 Mesoporous properties of the nanocomposites carbonized at different temperatures. Sample code
CoC1000–20
CoC800–20
CoC600–20
CoC400–20
C-800a
C-NO2b
SBET (m2/g) Vp (cm3/g)
59.4 0.139
102.1 0.143
166.7 0.251
155.4 0.175
0.6 0.012
135.2 0.114
a
Sample obtained by pyrolysis of the pure PAN microspheres. Sample obtained by pyrolysis of the product prepared under the same condition of CoC-800–20 microspheres, but without CoCl2. b
Table 3 Textural parameters of the nanocomposites carbonized under different heating rates.
SBET (m2/g) Vp (cm3/g) Vmicro (cm3/g)a a
CoC600–20
CoC600–40
CoC600–60
CoC800–20
CoC800–40
CoC800–60
166.7 0.251 0.042
180.4 0.341 0.060
200.1 0.291 0.078
102.1 0.143 0.014
167.7 0.135 0.063
232.5 0.269 0.077
Micropore volume of the nanocomposites calculated by t-plot method.
same time. It is probably because that under the rising heating rates, the removal of gaseous species was more intensely to form the narrower size pores, which also enlarged the surface areas. These results indicate that the pore structure can be tuned by changing the carbonization temperature and heating rates. Sample carbonized at 600 °C with a heating rate of 60 °C/min (CoC-600– 60) shows a specific surface area of 201 m2/g, which was chose as the absorbent in the methyl orange adsorption test.
3.2. Methyl orange adsorption on mesoporous magnetic nanocomposites The treatment of dye in wastewater is very important for the protection of environment [29]. With the large surface area and excellent magnetic properties, the mesoporous magnetic nanocomposites should become an excellent adsorbent for removal of pollutants in water by adsorption. As a model pollutant, MO was chosen as an adsorbate for the adsorption–desorption experiments. As shown in Fig. 4, the color of the solution changed from orange to colorless after addition of the mesoporous magnetic nanocomposites to the MO solution. The nanocomposites with adsorbed pollutant were quickly separated by magnet and the solution became colorless. By stirring in ethanol, the absorbed MO could be released from the nanocomposites easily.
Fig. 3. (a) Nitrogen adsorption–desorption isotherms and (b) corresponding pore size distributions of mesoporous magnetic nanocomposites carbonized at different temperatures, the heating rate was 20 °C/min.
14
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
Fig. 4. Photographs for removal of methyl orange (MO) in water by mesoporous magnetic nanocomposites CoC-600–60: (a) MO water solution; (b) CoC-600–60 was added into the solution to adsorb the MO; (c) CoC-600–60 was separated from water by magnetic field and the solution became colorless; (d) MO was released after CoC-600–60 with adsorbed MO was transferred into ethanol.
Fig. 5. (a) Equilibrium adsorption isotherms and (b) langmuir plot for MO adsorption by CoC-600–60.
Fig. 6. MO adsorption amounts of CoC-600–60 and CoC-1000–20 as a function of time.
Fig. 7. Relative adsorption capacity of CoC-600–60 for MO in water at room temperature for different cycles.
To evaluate the adsorption capacities of the as-obtained mesoporous magnetic nanocomposites for MO, the adsorption isotherm was conducted, as shown in Fig. 5. Various adsorption models have been proposed so far to study many kinds of adsorption systems. The adsorption behavior in our system has been successfully
described by Langmuir model. For this, the following Langmuir equation was employed [30]:
Qe ¼
Q mCe ðK d þ C e Þ
P. Zhang et al. / Journal of Colloid and Interface Science 389 (2013) 10–15
where Qm (mg/g) is the maximum adsorption capacity, and Kd is the dissociation constant. The plots of Ce/Qe against the equilibrium concentration Ce are shown in Fig. 5b. A good linear relationship is obtained for MO adsorption (R2 = 0.9962), which also confirms that the sorption followed the Langmuir model. As 1/Qm is the slope of the straight line in Fig. 5b, it can be estimated that the maximum MO adsorption capacity of CoC-600–60 is about 380 mg/g. Fig. 6 illustrates the MO adsorption amounts of CoC-600–60 and CoC-1000–20 as a function of time. It can be seen that the adsorption amount of MO increases with the time elongation and it is very quickly at the beginning and reached equilibrium for about 2 h, which demonstrates the fast adsorption rate of the magnetic porous nanocaomposites. Interestingly, the absorbed MO can be released almost completely from the mesoporous magnetic nanocomposites in ethanol. Due to the good regeneration ability and excellent magnetic separation properties, it is possible to quickly remove pollutants in water by adsorption–desorption cycle. Fig. 7 shows the recycling of CoC-600–60 nanocompaoites for the adsorption of MO in water. After five cycles, the mesoporous magnetic nanocomposites still retained the adsorption level over 85% for the pollutant.
4. Conclusions In summary, we have prepared magnetically separable mesoporous nanocomposites via the carbonization of PAN microspheres entrapped with cobalt precursors. During the carbonization process, the cobalt nanocrystals were formed and then protect by the in situ formed carbon from oxidation. The Co-NPs/ carbon nanocomposite exhibits high magnetization and low coercivity, and they can be easily separated from the solution by magnetic field. Besides, the Co-NPs/carbon nanocomposites have high surface area and a mesoporous structure, which can be tuned by controlling the carbonization temperature and heating rates. The synthesized mesoporous magnetic nanocomposites showed an excellent adsorbed ability to MO (380 mg/g), and the adsorbed MO can be easily released in ethanol. Therefore, it is expected to be a promising adsorbent for dyes in industrial wastewater.
15
Acknowledgments This work was supported by National Science Foundation of China (Grant nos. 20974023, 21034003, 51073040 and 21128001). References [1] E.J. Weber, R.L. Adams, Environ. Sci. Technol. 29 (1995) 1163–1170. [2] B. Royer, N.F. Cardoso, E.C. Lima, V.S.O. Ruiz, T.R. Macedo, C. Airoldi, J. Colloid Interf. Sci. 336 (2009) 398–405. [3] Z. Aksu, J. Yener, Waste Manage. 21 (2001) 695–702. [4] F. Herrera, A. Lopez, G. Mascolo, P. Albers, J. Kiwi, Appl. Catal. B: Environ. 29 (2001) 147–162. [5] A.W.M. Ip, J.P. Barford, G. McKay, J. Colloid Interf. Sci. 337 (2009) 32–38. [6] T. Otowa, Y. Nojima, T. Miyazaki, Carbon 35 (1997) 1315–1319. [7] R.-L. Tseng, F.-C. Wu, R.-S. Juang, Carbon 41 (2003) 487–495. [8] K. Mohanty, M. Jha, B.C. Meikap, M.N. Biswas, Ind. Eng. Chem. Res. 44 (2005) 4129–4138. [9] B. Özkaya, J. Hazard. Mater. 129 (2006) 158–163. [10] H.M. Chen, P.K. Chu, J.H. He, T. Hu, M.Q. Yang, J. Colloid Interf. Sci. 359 (2011) 68–74. [11] R. Afkhami, J. Hazard. Mater. 174 (2010) 398–403. [12] M. Fröba, R. Köhn, G. Bouffaud, Chem. Mater. 11 (1999) 2858–2865. [13] A.B. Fuertes, P. Tartaj, Chem. Mater. 18 (2006) 1675–1679. [14] D.W. Wang, F. Li, G.Q. Lu, H.-M. Cheng, Carbon 46 (2008) 1593–1599. [15] M. Sevilla, T. Valdés-Solís, P. Tartaj, A.B. Fuertes, J. Colloid Interf. Sci. 340 (2009) 230–236. [16] Y.F. Zhu, E. Kockrick, S. Kaskel, T. Ikoma, N. Hanagata, J. Phys. Chem. C 113 (2009) 5998–6002. [17] S.M. Zhu, D. Zhang, Z.X. Chen, Y.M. Zhang, J. Mater. Chem. 19 (2009) 7710– 7715. [18] J. Jang, H. Yoon, Small 1 (2005) 1195–1199. [19] Z.H. Sun, L.F. Wang, P.P. Liu, S.C. Wang, B. Sun, D.Z. Jiang, F.S. Xiao, Adv. Mater. 18 (2006) 1968–1971. [20] X.P. Dong, H.R. Chen, W.R. Zhao, X. Li, J.L. Shi, Chem. Mater. 19 (2007) 3484– 3490. [21] A.H. Lu, W. Schmidt, N. Matoussevitch, H. Bönnemann, B. Spliethoff, B. Tesche, E. Bill, W. Kiefer, F. Schüth, Nanoengineering of a magnetically separable hydrogenation catalyst, Angew. Chem., Int. Ed. 43 (2004) 4303–4306. [22] Y.W. Chu, P. Zhang, J.H. Hu, W.L. Yang, C.C. Wang, J. Phys. Chem. C 113 (2009) 4047–4052. [23] J. Jang, H. Yoon, Adv. Mater. 15 (2003) 2088–2091. [24] T. Yogo, S. Naka, J. Mater. Sci. 24 (1989) 2115–2119. [25] C.Q. Cui, S.P. Jiang, A.C.C. Tseung, J. Electrochem. Soc. 138 (1991) 94–100. [26] Y.Q. Wang, C.M. Yang, W. Schmidt, B. Spliethoff, E. Bill, F. Schüth, Adv. Mater. 17 (2005) 53–56. [27] V. Mayagoitia, F. Rojas, I. Kornhauser, Pore network interactions in ascending processes relative to capillary condensation, J. Chem. Soc., Faraday Trans. 81 (1985) 2931–2940. [28] J.S. Hummelshøj, R.Z. Sørensen, M.Y. Kustova, T. Johannessen, J.K. Nørskov, C.H. Christensen, J. Am. Chem. Soc. 128 (2006) 16–17. [29] S. Shin, H. Yoon, J. Jang, Catal. Commun. 10 (2008) 178–182. [30] I. Langmuir, J. Am. Chem. Soc. 40 (1918) 1361–1403.