Carbon Hybrids and Their

Jan 24, 2012 - The syntheses of SBA-15 hard silica template and the host structure of CMK-3 ...... Phosphorus removal and recovery from secondary effl...
0 downloads 0 Views 711KB Size
Article pubs.acs.org/Langmuir

Synthesis and Characterization of γ-Fe2O3/Carbon Hybrids and Their Application in Removal of Hexavalent Chromium Ions from Aqueous Solutions Maria Baikousi,† Athanassios B. Bourlinos,‡ Alexios Douvalis,‡ Thomas Bakas,‡ Dimitrios F. Anagnostopoulos,† Jiři Tuček,§ Klára Šafárǒ vá,§ Radek Zboril,§ and Michael A. Karakassides*,† †

Department of Materials Science and Engineering, ‡Physics Department, University of Ioannina, GR-45110 Ioannina, Greece Regional Centre of Advanced Technologies and Materials, Department of Physical Chemistry and Experimental Physics, Faculty of Science, Palacky University, Slechtitelu 11, 783 71 Olomouc, Czech Republic

§

ABSTRACT: Magnetic Fe2O3/carbon hybrids were prepared in a two-step process. First, acetic acid vapor interacted with iron cations dispersed on the surface of a nanocasted ordered mesoporous carbon (CMK-3). In the second step, the primarily created iron acetate species underwent pyrolysis and transformed to magnetic iron oxide nanoparticles. X-ray diffraction, Fouriertransform infrared, and Raman spectroscopies were used for the chemical and structural characterization of the hybrids, while surface area measurements, thermal analysis, and transmission electron microscopy were employed to determine their physical, surface, and textural properties. These results revealed the preservation of the host carbon structure, which was homogenously and controllably loaded (up to 27 wt %) with nanosized (ca. 20 nm) iron oxides inside the mesoporous system. Mössbauer spectroscopy and magnetic measurements at low temperatures confirmed the formation of γ-Fe2O3 nanoparticles exhibiting superparamagnetic behavior. The kinetic studies showed a rapid removal of Cr(VI) ions from the aqueous solutions in the presence of these magnetic mesoporous hybrids and a considerably increased adsorption capacity per unit mass of sorbent in comparison to that of pristine CMK-3 carbon. The results also indicate highly pH-dependent sorption efficiency of the hybrids, whereas their kinetics was described by a pseudo-second-order kinetic model. Taking into account the simplicity of the synthetic procedure and possibility of magnetic separation of hybrids with immobilized pollutant, the developed mesoporous nanomaterials have quite real potential for applications in water treatment technologies.



INTRODUCTION Ordered mesoporous carbons of CMK-n (n = 1−9) type are unique porous carbonaceous materials that possess a wellordered pore structure with large surface area and high specific pore volume, along with tunable pore size.1,2 These materials, which have been synthesized using well-ordered hexagonal and cubic mesoporous silicate materials as hard templates and sucrose as the carbon source,3,4 have gained considerable interest as new nanostructured materials due to their potential applications as adsorbents, catalyst supports, and materials for advanced electronics applications.5−10 One of the most highly investigated carbon materials of this family is CMK-3, which is synthesized by applying the mesoporous silica sieve SBA-15 as © 2012 American Chemical Society

a template. Its structure is composed from uniformly sized carbon nanorods, arranged in a hexagonal pattern and connected by means of carbon spacers. Due to its open pore structure and high surface area, CMK-3 provides marked advantages over common activated carbon in the adsorption and diffusion processes.11,12 On the other hand, nanoscale magnetic iron oxides offer great potential as oxidatively stable magnetic particles with diverse applications in electronics, optoelectronics, medicine, Received: October 13, 2011 Revised: January 18, 2012 Published: January 24, 2012 3918

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

magnetic storage, and biotechnologies.13−20 These nanosized materials display properties that differ from their corresponding bulk material counterparts. For instance, iron and iron oxide magnetic nanoparticles have received considerable attention concerning the cleanup of environmental contaminants because of their small particle size, high surface area, catalytic activity, low cost, and ease of preparation.21−25 However, iron nanoparticles are sometimes unstable, and their stabilization particularly in terms of aggregation and oxidation in air is a crucial point to be solved. Immobilization and stabilization of these nanoparticles can be achieved when they are embedded into a host matrix. Thus, the materials could be protected against oxidation and could be randomly dispersed without minimizing their ability for functionalization.17 As the porous support may additionally affect the catalytic and sorption properties of the hosted nanoparticles, it is expected that a suitable combination of nanoparticles and host could generate hybrids with novel properties that are not displayed in the pristine materials. Along these lines, by combining the CMK-3 with magnetic iron oxide nanoparticles it is possible to develop new carbon hybrids endowed with improved properties through synergetic effects. The porosity of the CMK-3 matrix may induce accessibility of the active catalytic particles, affecting their selectivity and activity. From the environmental point of view, such magnetic hybrids could be used as adsorbents with a high effectiveness because of their enhanced ability for adsorption and/or concurrent reduction of multiple types of pollutants. As a superior aspect, these materials with immobilized pollutants can be easily separated and collected by an external magnetic field. Until now, numerous published research works have focused on the dispersion of magnetic iron oxide nanoparticles into CMK-3 porous matrices.26,27 In particular, a subset of these research works is concerned with the synthesis and characterization of mesoporous carbon/iron oxides composites but have not reported on the final developed hybrids for catalytic or environmental applications. Moreover, among these, the synthetic route reported involves impregnation with iron nitrate salts leading, upon heating in air, to undesired properties of the iron nanoparticles, for instance, the formation of nonmagnetic iron nanoparticles, such as hematite,27 or mixtures of magnetic/nonmagnetic iron oxides or iron carbides phases.26,28,29 In contrast, recently reported one-pot synthesis of mesoporous/iron oxides hybrids30 has been based on a specialized coassembly process using resol as carbon source, ferric citrate as magnetic precursor, and triblock copolymer P127 as template, which yielded a single phase of γ-Fe2O3 but with the disadvantage of lower specific surface area (1000 m2/g). Along these lines, γ-Fe2O3/carbon composites with γ-Fe2O3 nanoparticles have been synthesized by a co-casting method31 using wetness impregnation of FeCl3·6H2O into a polymerizing furfuryl alcohol/SBA-15 composite and thermal treatment at 873 K in Ar atmosphere. However, the method yields iron nanoparticles mainly incorporated but not dispersed into the carbon rods and, thus, nonaccessible or chemically active (surface area data). On the other hand, few studies have been reported concerning the use of magnetic OMCs for catalytic and environmental applications. For instance, CMK-3 carbons with (a) dispersed Fe3O4 nanoparticles have been synthesized and

tested as adsorbents and/or catalysts for dry oxidation of phenol;32 (b) Fe3O4 nanoparticles, α-Fe, and FeC have been synthesized according to ref 28 and used for adsorption of fluorescein, a fluorescence dye from aqueous solutions;33 and (c) dispersed ferrous and ferric ions cross-link with various functional groups of carbon matrix have been prepared and are used for arsenic adsorption.34 Although charcoal and activated carbons have been considered as potential adsorbents for removal and recovery of chromium ions from aqueous solutions,35 until now none of these studies have evaluated these carbon/iron oxide systems in environmental remediation processes, such as the adsorption of chromium from aqueous solutions. In the present work we prepared magnetic hybrids from iron oxide nanoparticles and CMK-3 carbon, using hard template synthesis, in conjunction with the ability of iron carboxylate compounds to afford upon pyrolysis crystalline magnetic phases.36−38 These hybrids have the crucial advantage of exhibiting controllable loading of magnetic γ-Fe2O3 nanoparticles while retaining at the same time its porous carbon matrix for efficient adsorption kinetic mechanisms, also offering high specific surface area. The final hybrid materials were comparatively tested, together with pristine CMK-3 carbons, for their ability to separate and immobilize Cr(VI) ions from aqueous solutions. Chromium is a heavy metal widely used in many engineering and chemical industries due to its durability and aesthetic quality.39 Unfortunately, it is considered a major pollutant, since in the oxidation state of Cr(VI) it is toxic and carcinogenic to humans and animals. Furthermore, Cr(VI) is highly soluble and mobile and exerts toxic effects on biological systems due to its oxidizing properties.40,41 Here, the Cr(VI) reaction kinetics and adsorption isotherms were performed to demonstrate the efficiency of the new hybrids. The effects of pH and concentration of pollutant on kinetics have been also evaluated.

2. EXPERIMENTAL SECTION 2.1. Chemicals. All chemical reagents were used as purchased without further purification. Triblock copolymer P123 (EO20PO70EO20, EO = ethylene oxide, PO = propylene oxide), tetraethylorthosilicate (TEOS) 98%, sucrose, iron, nitrate nonahydrate, acetic acid 99.7%, potassium dichromate, phosphoric acid 85%, and 1,5-diphenylcarbazide were purchased from Sigma-Aldrich. Hydrochloric acid 37%, sulfuric acid 95−97%, sodium hydroxide, nitric acid (HNO3) 65%, and sodium chloride (NaCl) were purchased from Merck, whereas ethanol (EtOH) 99.5%, methanol 99.8%, and acetone 99.95% were purchased from Panreac and Fischer Scientific, respectively. 2.2. Synthesis of SBA-15 Silica and CMK-3 Carbon. The syntheses of SBA-15 hard silica template and the host structure of CMK-3 carbon were performed according to the procedures reported elsewhere.42,43 In the final stage of synthesis the mesoporous carbon CMK-3 was obtained by removing the silica framework after treatment twice with 1 M solution of NaOH in ethanol−water (v/v = 1/1) at 80 °C. Oxidized mesoporous carbon (CMK-3-O) was also used as a support for the magnetic iron oxide nanoparticles. To this aim, CMK-3 was chemically treated with HNO3 as follows: 0.2 g of CMK-3 was dispersed in 10 mL of aqueous solution HNO3 (10%) and the mixture was refluxed at 110 °C for 1 h. The mixture was centrifuged, washed with H2O several times (until pH = 7), and dried at 90 °C for 12 h. 2.3. Synthesis of Hybrid Materials. The high surface area of CMK-3 and of its oxidized form CMK-3-O makes them ideal matrices for the deposition of magnetic nanoparticles toward the corresponding hybrid materials. The hybrid materials CMK-3@mx (x = 1, 2, 10) and 3919

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

X-ray powder diffraction data were collected on a D8 Advance Bruker diffractometer using Cu Kα (40 kV, 40 mA, λ = 1.541 78 Å) radiation and a secondary beam graphite monochromator. Diffraction patterns were collected in the 2θ range 0.5−80°, in steps of 0.02°, with 2 s counting time per step. Thermogravimetric (TGA) and differential thermal (DTA) analysis were performed using a Perkin-Elmer Pyris Diamond TG/DTA. Samples of approximately 5 mg were heated in air from 25 to 900 °C, at a rate of 5 °C/min. UV−visible (UV−vis) spectra of solutions were measured in quartz vesicle with a UV2401(PC)-Shimadzu two beam spectrophotometer in the range 400− 700 nm at a step of 0.5 nm using a halogen lamp. Transmission electron microscopy (TEM) observations were performed on a JEOL JEM-2010 transmission electron microscope equipped by LaB6 cathode (accelerating voltage of 200 kV; point-to-point resolution of 0.194 nm). A drop of high-purity distilled water, containing the ultrasonically dispersed particles, was placed onto a holey carbon film supported by a copper-mesh TEM grid and air-dried at room temperature. High resolution X-ray fluorescence spectroscopy (XRF) has been applied to measure the X-ray emission spectra of the elements in the samples. The spectra are measured with a wavelengthdispersive (WD-XRF) X-ray crystal spectrometer (S4 Explorer, Bruker). The spectrometer is equipped with a 1 kW Rh tube for target excitation, three different crystal analyzers (XS-55 multilayer, PET(002) and LiF(200) with 2d-spacing of 55, 8.75, and 4.03 E, respectively) covering all the periodic table elements from C on, and two detectors, a flow gas counter for the low photon energies and a scintillation counter for the high energy X-rays. The N2 adsorption−desorption isotherms were measured at 77 K on a Sorptomatic 1990, Thermo Finnigan porosimeter. All samples used for the surface analyses were outgassed at 120 °C for 20 h under vacuum (10−4 mbar) before the measurements. Specific surface areas SBET were determined with the Brunauer−Emmett−Teller (BET) method using adsorption data points in the relative pressure P/Po range of 0.01−0.30. Pore size distributions of mesoporous pristine and hybrid carbons weres calculated using the Barrett−Joyner−Halenda (BJH) model. The total pore volume was estimated from the adsorbed amount at relative pressure of 0.99, whereas the external surface area (Sext) mesoporous surface area (Smeso) and micropore volume (Vmicro) were determined using V−t-plot method. Mössbauer spectra were recorded with a conventional constant acceleration spectrometer and a 57Co(Rh) source and the parameters were obtained by a least-squares minimization program assuming Lorentzian line shapes. The Mössbauer spectrometer was calibrated with an α-Fe absorber and all isomer shift values reported here are relative to iron at room temperature. Magnetic measurements were carried out at room temperature in a Quantum Design Magnetometer (SQUID), while the samples were measured in their powder form. Thermal variation of zero field cooled (ZFC) and field cooled (FC) magnetization curves were taken from 250 to 4.2 K.

CMK-3-O@m4 consist of mesoporous carbon CMK-3 or oxidized mesoporous carbon CMK-3-O, respectively, and magnetic iron oxide nanoparticles in four different loadings. The numbers 1, 2, 10, and 4 in code names CMK-3@mx (x = 1, 2, 10) and CMK-3-O@m4 denote the initial mass ratio Fe(NO3 )·9H2 O/CMK-3 or CMK-3-O, respectively. More specifically, 100 mg of CMK-3 was wetimpregnated with 5 mL of iron nitrate solution in methanol (100 mg Fe(NO3)·9H2O in 5 mL methanol) and then the solvent was rapidly removed at 80 °C. The obtained solid was powdered and then exposed to vapors of acetic acid at 80 °C for 1 h. The solid powder was dried for 15 min at 80 °C to remove any physically absorbed acetic acid. The hybrid magnetic material CMK-3@m1 was obtained after pyrolysis for 20 min in Ar flowing atmosphere at 400 °C. Furthermore, CMK-3@m2 and CMK-3@m10 hybrid materials were also prepared in a similar manner but using 200 mg iron nitrate or 1 g iron nitrate, respectively. In case of the hybrid material CMK-3-O@m4, 100 mg of CMK-3-O was wet impregnated by filtration with 20 mL of iron nitrate solution in methanol (400 mg Fe(NO3)·9H2O in 20 mL methanol) and then the same procedure was followed. 2.4. Batch Experiments. The pristine mesoporous carbon CMK-3 and the hybrid magnetic material CMK-3@m1 have been used to remove Cr(VI) from aqueous solutions. Eighteen milligrams of CMK3 or CMK-3@m1 was added to 100 mL solution of K2Cr2O7 in distilled water (three different concentrations Cr(VI): 0.6 mg/L, 2 mg/L, and 6 mg/L) and stirred at room temperature for 24 h in a closed vessel. The pH of the suspensions was adjusted to 5.5 or 3 by adding hydrochloric acid solution (1 N). During the reaction, at different periods (0 h, 0.5 h, 1 h, 2 h, 3 h, 6 h, 9 h, and 24 h), 5 mL of the under study suspension was withdrawn and centrifuged. From this overnatant solution, 0.3 mL of the Cr(VI) solution was used for colorimetric measurements, while the solid and the remaining centrifuged solution were returned to the suspension. The Cr(VI) concentrations of the solutions were determined by using the 1,5diphenylcarbazide method.44 In addition, because pH has a strong effect on Cr(VI) sorption and reduction, the pH value of the suspension was also measured at the different periods. The 1,5diphenylcarbazide method is based on the reaction of Cr(VI) cations with 1,5-diphenylcarbazide molecules leading to the formation of a red−purple chromium 1,5-diphenylcarbazide complex. In particular, in case of initial concentration of 0.6 mg/L, 3 mL of the Cr(VI) solution at the different periods was mixed with 120 μL 1,5-diphenylcarbazide (0.025 g) solution in acetone (10 mL) and 60 μL H3PO4 solution (0.5 mL H3PO4 (85%) in 10 mL of H2O). The solution was left for 10 min to allow color development, and then the solution’s concentration was determined spectrophotometrically at 540 nm using a calibration curve. It is remarkable to note that in the case of initial concentration of 2 mg/L or 6 mg/L the measurements have been done after making appropriate dilutions in distilled water. In particular, in these cases 0.3 mL of the Cr(VI) solution at the different periods was mixed with 2.7 mL distilled water, 120 μL 1,5-diphenylcarbazide solution, and 60 μL H3PO4 solution. The Cr(VI) concentration of the solutions, in these cases, was determined in consideration of these dilutions. At the end of the experiment, the suspension was centrifuged and the residual solids were washed 2 times with 50 mL of deionized water and subsequently dried in air. The residual solids after the batch experiments were analyzed by WXRF. The calibration curve was created from different standard solutions of Cr(VI) (known concentrations 0−1 mg/L), and it was determined that absorbance at 540 nm exhibits a linear relationship with the Cr(VI) concentration (Absorbance = 0.85186 × [Cr(VI)] − 0.00836, R2 = 0,998). The pHpzc (points of zero charge) of CMK-3 and CMK-3-O were determined using the pH drift method45 with the modification that NaCl was used as an inert electrolyte. 2.5. Characterization. Infrared (FT-IR) spectra of samples in powder form, dispersed in KBr pellets, which were the average of 32 scans at 2 cm−1 resolution, were measured with a Perkin-Elmer GX, Fourier transform spectrometer in the frequency range 400−4000 cm−1. Raman spectra were recorded with a micro-Raman system RM 1000 Renishaw using a laser excitation line at 532 nm (Nd:YAG) in the range 500−2000 cm−1. Laser power of ∼10mW was used with 2 μm focus spot in order to avoid photodecomposition of the samples.

3. RESULTS AND DISCUSSION 3.1. Structural Characterization. The infrared spectra of CMK-3, CMK-3-O@m4, and the materials prepared in the intermediate stages using CMK-3-O as support are shown in Figure 1. The spectrum of CMK-3 (a) shows a broad absorption envelope in the frequency region of 1700−1000 cm−1, exhibiting two maxima at around 1580 cm−1 and 1165 cm−1. According to previous studies on graphitized carbon46 these peaks can be assigned to the stretching vibrations of C C and C−H bonds in aromatic carbon rings, respectively. On the other hand, the spectrum of CMK-3-O sample (b) is indicative of oxidation of the carbon surface after its chemical treatment with nitric acid. For instance, the increase in the absorptions at 1720 cm−1 and 1590 cm−1 can be assigned to asymmetric stretching vibrations of the newly formed −COOH carboxyl and −COO− carbonyl and/or −CO ketone units.47−51 The band at 1340 cm−1 can be attributed to symmetric stretching vibrations of −COOH groups, whereas 3920

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

formation of iron nanoparticles on carbon surfaces is also indicative of the spectrum of CMK-3-O@m4 hybrid. For instance, in the low frequency region (745−500 cm−1) a new broad band appears that probably originates from Fe−O vibrations of the formed iron oxide nanoparticles. In fact, magnetite (Fe3O4) and maghemite (γ-Fe2O3) exhibit bands at 570 cm−1 and 630 cm−1, respectively, which can be assigned to the Fe−O stretching modes of the tetrahedral and octahedral sites in their inverse spinel structure. Thus, the formation of such iron oxide phases can be clearly suggested. The spectra of the CMK-3@mx (x = 1, 2, and 10) hybrids showed also similar spectral characteristics with more or less differences in the absorption intensities in the low frequency region due to their different iron loading. Raman spectra of the pristine CMK-3 and CMK-3-O@m4 and CMK-3@mx (x = 1, 2, and 10) hybrids are shown in Figure 2. The spectra of all samples exhibit two intense peaks at

Figure 1. FT-Infrared spectra of samples: CMK-3 (a), CMK-3-O (b), and CMK-3-O after its wet impregnation with iron nitrated solution (c), and then after exposure to vapors of acetic acid (d), and magnetic hybrid CMK-3-O@m4 (e) prepared by pyrolysis of sample d (for the designations see the text).

the band at 1230 cm−1 to asymmetric stretch of −C−C−C bridges in ketonic groups and/or to deformation vibrations of O−H in the carboxylic acid groups. Also, the weak band at 665 cm−1 can be assigned to bending O−C−O vibrations of carboxyl units, which overall seems to exist also in a small concentration in parent CMK-3. In Figure 1 the infrared spectrum of CMK-3-O (c) after its wet impregnation with iron nitrate solution is also shown. The presence of nitrate anions, which remain in mesoporous carbon after the drying, is demonstrated by the appearance of a sharp peak at 1382 cm−1, and another one weaker in intensity at 830 cm−1. Both peaks are attributed to asymmetric and bending vibrations of NO3− anions, respectively. In addition, the remarkable increase in the absorption at 670 cm−1 could be assigned to the development of iron-carboxylates in various configurations. More drastic changes in the infrared spectrum (d) were the result of the exposure of this sample (c) to vapors of acetic acid during the next step of synthesis. In fact, the significant increase in the absorptions at 1590 and 661 cm−1 as well as the new bands at 1445 and 613 cm−1 can all be assigned to vibrations of a formed trinuclear Fe(III)-acetato complex. Specifically the strong bands at 1590 cm−1 and 1445 cm−1 are assigned to asymmetric and symmetric vibrations of COO− bridged to Fe(III) and the weaker bands at 661 cm−1 and 613 cm−1 to OCO and COO deformations,38,52 respectively. These vibrations are not observed in the infrared spectrum of CMK-3-O@m4 hybrid (Figure 1e) after the pyrolysis of sample (d), indicating that the acetate ligands have been removed from the coordination sphere of iron ions. On the contrary, the basic absorption peaks at 1580 cm−1 and 1230 cm−1, due to vibrations of carbonyl, ketone units, and −C−C−C bridges are still observed, while the band at 1700 cm−1, which was assigned to carboxyl groups, has notably been reduced in intensity. Thus, after all synthetic steps the surface of CMK-3-O@m4 hybrid seems to remain similar to that of the parent CMK-3 regarding the basic carbon structure and the main active groups on it. On the other hand,

Figure 2. Raman spectra of CMK-3, CMK-3-O@m4, and CMK-3@ mx (x = 1, 2, and 10) samples. Inset shows the low frequency spectrum of γ-Fe2O3 nanoparticles recorded from a reference sample.

1598 cm−1 and 1335 cm−1 which are indicative of the sp2 and sp3-hybridized carbon atoms, respectively. Specifically, the first band at 1598 cm−1 is assigned to the E2g tangential mode (Gband, see Figure2, inset), while the second band at 1335 cm−1 (D-band) originates from defects which break the basic symmetry of the graphene sheet of carbon rods. It is interesting to note that the G-band peak in all Raman spectra is located at a higher frequency than that in graphite (1598 vs 1580 cm−1). This blue shift of the G-bands in CMK-3 samples can be understood as a result of overlapping between the G-band and the new scattering peak at around 1620 cm−1 (D′-band). The latter band is activated in the case where a graphitized carbon structure has a sufficient concentration of defects. In addition, a possible explanation involves the scattering contribution from a resonance, originating from isolated double bonds or from an alternative pattern of single−double carbon bonds in the samples.53 It is also possible to assess the graphitization degree of carbon rods of CMK-3 and CMK hybrids from the intensity ratio (ID/IG) and the full width at half-maximum (fwhm) of the G-band. For instance, these parameters for all samples have 3921

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

Figure 3. X-ray diffraction patterns of CMK-3, CMK-3-O, CMK-3-O@m4, and CMK-3@mx (x = 1, 2, and 10) samples at low (left) and high (right) scattering angles.

Figure 4. DTA (left) and TGA (right) curves of CMK-3 and CMK-3-O@m4 samples.

carbon (CMK-3-O), and magnetic carbon hybrids (CMK-3@ mx, x = 1, 2, and 10 and CMK-3-O@m4) prepared using different weight ratios (x = 1, 2, 4, 10) of iron-source/CMK-3. The XRD patterns of CMK-3 samples and the patterns of the samples subjected to oxidation treatment (CMK-3-O) (Figure 1a and b) all exhibit three peaks in the range 2θ = 0.75−3ο which can be indexed to the (100), (110), and (200) reflections of the hexagonal space group (P6mm). This indicates a highly ordered arrangement of the interconnected carbon nanorods in the parent materials. On the other hand, the patterns of magnetic carbon hybrids show a small shift to smaller d-spacing for the (100) reflection and a remarkable decrease in the intensity of all reflections, the decrease being dependent on the specific iron oxide loading. Such behavior does not necessarily suggest the partial or complete damage of the structural order of CMK-3@mx solids,

similar values; i.e., the ratio (ID/IG) is in the range 0.8−0.95 and the fwhm is ca. 110 cm−1. These values are typical for CMK-3 structures,54 indicating a rather low ordering (bulk graphite has ID/IG = 0.1−0.3 and fwhm ∼20 cm−1) due to the threedimensional structure of the carbon rods. On the other hand, the development of iron nanoparticles affects the scattering intensity in the region 800−400 cm−1 where a broad band appears for high iron oxide concentration (CMK-3@m10, Figure 2). In the same region the magnetic iron oxide phases exhibit Raman bands, as clearly illustrated in the spectrum (γFe2O3) which is recorded from a reference sample containing small nanoparticles of maghemite. Overall these findings suggest that the graphitized structure of carbon remains unimpaired and the nanoparticle formation is also well-proven. Figure 3, left, shows the powder XRD patterns from the pristine mesoporous carbon (CMK-3), activated mesoporous 3922

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

Figure 5. N2 adsorption−desorption isotherms (left) and representative t-plots (right) from CMK-3, CMK-3-O, CMK-3-O@m4, and CMK-3@m1 samples.

place at a lower temperature. The same phenomenon has also been observed concerning the combustion temperature of the other three hybrids (358/361/365 °C). This behavior should be assigned to the existence of small magnetic nanoparticles on the surfaces of hybrids which act as catalysts for the carbon combustion since their exothermic oxidation takes place at lower temperatures.59 Moreover, the TGA curve of the CMK-3 sample showed a steep weight loss of nearly 100%, between 250 and 450 °C, which is also attributed to the combustion of carbon, while the analogous TGA curve of CMK-3-O@m4 hybrid showed a weight loss of about 87.5%, indicating that the developed nanoparticles were 12.6% of the total mass of this specific hybrid. Based on the TGA curves (not shown here) of the other hybrids, the iron oxide content (Fe2O3) has been calculated to be in the range from 11.5 wt % (CMK-3@m1) to 27.3 wt % (CMK-3@m10). 3.2. Textural, Surface, Morphological, and Magnetic Properties. The N2 adsorption−desorption isotherms for the CMK-3, CMK-3-O carbons and CMK-3@m1, CMK-3-O@m4 hybrids are shown in Figure 5. All samples exhibit typical isotherms of ordered mesostructures (type IV) with a welldefined step in the adsorption curve, near a 0.5 value of P/Po, indicating filling of the framework-confined uniform mesopores. The parent CMK-3 carbon reveals a high specific surface (SBET) of 1650 m2/g, while from the pore size distribution curve, the average pore diameter (dBJH) was calculated to be ∼3.4 nm. The adsorption data of CMK-3 carbon were also treated with the Va−t plot method. In this case, a total surface area St of 1520 m2/g and a macropore area from textural mesopores, Sext of 151 m2/g were calculated using the slopes of Va−t plot. For the named calculation, slopes at special regions, referenced as (1) and (2) in Figure 5, right, were used, which refer to the relatively lower (0.6 < t) and higher (1 < t < 2) pressure range before and after the filling of mesopores, respectively. The mesopore surface area was calculated from St by subtraction of Smacro and was found to be ∼1369 m2/g. Moreover, the straight line extending from the lower pressure branch also gave a volume of nitrogen adsorbed inside the micropores of 0.05 cm3/g (intercept with y-axis). Similarly,

but probably can be correlated to some density effects originating from the different material constituting the walls (carbon) and the material filling the pore regions (iron oxide) of hybrids. In fact, a similar effect has already been discussed in previous studies dealing with metal oxide incorporation into the pores of silicate and carbon mesostructures.55−58 According to these studies the loss in intensity was due to the fact that the introduction of scattering material into the pores leads to an increased phase cancellation between scattering form walls and the pore regions. Figure 3, right, presents the powder XRD patterns within the 2θ range 28−44°, of CMK-3@mx, (x = 1, 2, 10) and CMK-3O@m4 samples along with parent CMK-3. The positions of the diffraction lines of γ-Fe2O3 and magnetite (Fe3O4) are also indicated. As shown in this figure, the patterns of CMK-3@m1 and CMK-3@m2 exhibit a broad reflection near 35.7° that is weak in intensity but distinguishable from the rest of the amorphous carbon pattern. Besides the increase of the weight ratio of iron-source/CMK-3, that peak becomes more intense and sharper, an effect which indicates that the size of the formed nanoparticles has also increased. That peak can now be indexed to the (311) Bragg reflection of γ-Fe2O3 or Fe3O4 [JCPDS no 24−008 and 86−1338]. In addition, the pattern of these samples exhibits additional reflections at about 30° and 43°, which can be attributed to the (220) and (400) lattice planes of the same oxide phases, respectively. From the full width at half-maximum of the (311) diffraction peak, obtained after curve fitting of the CMK-3@m patterns within the 2θ range 10−30°, the average sizes of the crystallites were estimated to be 20 nm (CMK-3@m10), 13 nm (CMK-3-O@ m4), and 8 nm (CMK-3@m2) using Scherrer’s equation, respectively. The as-calculated nanoparticle sizes seem to be in good agreement with those calculated from TEM images, as will be discussed further below. Figure 4 shows the DTA (left) and TG (right) curves for the parent CMK-3 and CMK-3-O@m4 hybrid. Both samples show very sharp exothermic peaks at 400 °C (CMK-3) and 356 °C (CMK-3-O@m4). These peaks can be attributed to carbon matrix combustion, which in the case of the hybrids has taken 3923

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

Table 1. Textural Properties of the CMK-3 Mesoporous Carbon and Magnetic Hybrids samples

SBETa (m2.g−1)

Stb (m2.g−1)

Smacrob (m2.g−1)

Smesob(m2.g−1)

Vmesob (cm3.g−1)

Vμb (cm3.g−1)

dBJHc (nm)

CMK-3 CMK-3-O CMK-3@m1 CMK-3-O@m4

1650 1248 1043 774

1520 1093 953 761

151 74 85 56

1369 1019 868 705

1.20 0.87 0.76 0.50

0.05 -

3.4 3.4 3.4 3.4

a

Total specific surface area form multipoint BET analysis. bTotal (St), macro (Smacrp), and mesoporous (Smeso) surface area; mesoporous (Vmeso) and micropore (Vμ) volume determined from Va−t-plots (standard: de Boer). cMean pore diameters determined form BJH analysis of the desorption data.

Figure 6. TEM images of CMK-3@m1 (a,b) and CMK-3-O@m4 (c,d) hybrids.

using the same analytical procedure to other points of the data curves, it has been possible to identify the mesopore volume from the higher pressure branch of the same curve (Vmeso = 1.2 cm3/g). All the above textural properties are summarized in Table 1. The CMK-3-O carbon also exhibits high surface area (1248 m2/g), but it seems that the activation treatments lead to a decrease of almost all textural parameters (Table 1). This reduction in adsorption capacity should be attributed to the chemically attached functional groups (i.e., −COOH) at the carbon surface, which cause blocking of the micropores (Vμ ≈ 0, Table 1) and decrease the available mesopore surface. On the other hand, the total surface area and pore volume decrease remarkably with the increasing iron oxide loading in hybrid samples (Table 1), which indicates the presence of some mesopore blocking by the nanoparticles. However, high surface areas were maintained even after iron oxide nanoparticle formation (1043 m2/g for CMK-3@m1 and 774 m2/g for CMK-3-O@m4 compared to 1650 m2/g and 1248 m2/g for CMK-3 and CMK-3-O, respectively). Transmission electron microscopy (TEM) has been used to study the quality of mesostructured carbon as well as the embedding and morphological characteristics of the related iron oxide nanoparticles. Figure 6a,b shows the TEM lattice images of the CMK-3@m1 sample. It can be observed in this figure that the hybrid exhibits a linear mesoporous array (white lines) separated by carbon rods (black lines), both of whose diameters

were estimated to be 2.4 and 7 nm, respectively. Using these values, the hexagonal unit cell ao parameter of CMK-3@m1 is estimated at ca. 9.4 nm. It should be noted that the pore diameter, as estimated from these images, is 2.4 nm and seems to be smaller than the diameter estimated by physisorption BJH calculations, dBJH = 3.4 nm. This difference probably originates from the fact that in the TEM images only 70% of the formed pore cylinders are observed, since the remaining part is overlapping in a visual context with carbon rods and is, thus, invisible in the pore calculations. On the other hand, the dark spots in the image, which overspread to pore channels and carbon rods, indicate nanoparticle formation (Figure 6a,b). The observed nanoparticles appear elliptical in shape and with a maximum size of about 20 nm. Figure 6c,d shows the TEM images of the CMK-3-O@m4 sample. As shown in these figures, rhombic or polygonal shaped iron oxide nanoparticles are homogeneously dispersed on the surfaces of the worm-like carbon particles of CMK-3. These nanoparticles, which are observed to be wedged partly within the mesoporous channels, exhibit sizes between 15 and 20 nm. All of the above results are consistent with the results derived from the powder XRD patterns and surface area measurements concerning the nanoparticle sizes and the ordering and/or porosity of the carbon structure. Characteristic Mössbauer spectra (MS) of samples CMK-3@ m1 and CMK-3-O@m4 collected at 300 K are shown in Figure 7. The spectra exhibit both quadrupole and magnetically split 3924

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

contribution in addition to a linear paramagnetic increase of M with respect to H is observed, which is shown better in the inserts of Figure 8a. Above 5 kOe a strong linear diamagnetic contribution with negative dM/dH slope is evident. The maximum absolute magnetization values are not exceeding 1.00 emu/g and no coercivity is observed. The features of the loop’s center at 300 K characterize the superparamagnetic state of the maghemite nanoparticles of the hybrid material, which at this temperature show very fast superparamagnetic relaxation. As a consequence, the magnetic moment corresponding to the nanoparticle assembly is quite low, allowing the diamagnetic contribution of the CMK matrix to appear and dominate the M values at fields greater than 5 kOe. At 5 K the situation is different. The loop shows ferromagnetic characteristics, with coercivity values on the order of 500 Oe, which indicate that the maghemite nanoparticles are magnetically blocked at this temperature.62−64 The M values are not saturated at high fields, indicating the presence of superparamagnetic relaxation in some part of the nanoparticle assembly, most probably that corresponding to the smallest nanoparticles. Lack of interparticle interactions can also be suggested to explain this behavior. The reduced values of the absolute maximum magnetizations at ±70 kOe (∼2.12 emu/g, Table 3) is caused by both superparamagnetic finite-size effects and the presence of the diamagnetic CMK matrix, which accounts also for a large part of the weighted mass of the sample. At this temperature the magnetic blocking of the maghemite nanoparticles is enough to suppress the appearance of the diamagnetic contribution of the CMK matrix, which is observed only at the 300 K loop where these nanoparticles show complete superparamagnetic behavior. The M versus temperature (T) measurements, shown in Figure 8, confirms further the superparamagnetic behavior of the maghemite nanoparticles. In particular, from Figure 8c it is evident that an applied field of 100 Oe is not strong enough to overcome the very fast superparamagnetic relaxation of the nanoparticle assembly, in order to align the magnetic moments of the nanoparticles to its direction. The decrease of the M values with increasing temperature both in the ZFC and FC modes denote lack of interparticle magnetic interactions between the nanoparticles. When the measurements are done under an applied field of 1000 Oe the picture is different. This external field is strong enough to disclose a magnetic blocking temperature of TB ≈ 50 K, as evident by the maximum of the ZFC curve. Moreover, the lack of saturation in the M values of the FC branch at low temperatures confirms further the suggestion for the absence of interparticle magnetic interactions.

Figure 7. Mössbauer spectra of CMK-3@m1 (a) and CMK-3-O@m4 (b) hybrids recorded at 300 K.

contributions; however, the former contribution dominates the absorption area in all spectra. At 300 K, the resonant absorption lines of the magnetically split contributions are rather broad, whereas a set of two magnetically (M1 and M2) and one quadrupole (SP) split components were used to fit these spectra. The Mössbauer hyperfine parameter values resulting from these fits are listed in Table 2. The isomer shift (IS), quadrupole splitting (QS), quadrupole shift (2ε), and hyperfine magnetic field (Bhf) values of all components characterize Fe3+ high spin (S = 5/2) ions in oxygen coordinated environment. Combining the information available from XRD and TEM measurements with the detection of only Fe3+ states with Mössbauer hyperfine parameters values, which correspond to ferric ions embedded in oxide spinel structure,60,61 we can conclude that the iron bearing phase in our samples is that of maghemite (γ-Fe2O3). Due to the particle size distribution, maghemite nanoparticles give both superparamagnetic and ferromagnetic Mössbauer signals at room temperature. Isothermal magnetization (M) measurements as a function of the applied magnetic field (H) for CMK-3@m1 collected at 300 and 5 K are shown in Figure 8. At 300 K the hysteresis loop is composed of two different contributions with opposite behavior. At low fields, up to ±5 kOe, a weak ferromagnetic

Table 2. Mössbauer Parameters Resulting from Least Square Fits of the 300 K Spectra of CMK-3@m1 and CMK-3-O@m4 Hybridsa T (K)

IS (mm/s)

Γ/2 (mm/s)

QS (mm/s)

Bhf (kOe)

ΔBhf (kOe)

A (%)

component

CMK-3@m1

300

CMK-3-O@m4

300

0.34 0.34 0.34 0.34 0.34 0.37

0.32 0.25 0.31 0.25 0.25 0.28

0.02 −0.02 0.70 0.02 0.02 0.78

483 403 0 491 353 0

9 160 0 13 160 0

15 47 38 18 30 52

M1 M2 SP M1 M2 SP

Sample

T, measurement temperature; IS, isomer shift; Γ/2, linewidth; QS, quadrupole splitting; Bhf, hyperfine magnetic field; ΔBhf, spreading of hyperfine magnetic field; A, relative spectral areas of individual spectral components.

a

3925

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

Figure 8. Magnetic hysteresis loops of the CMK-3@m1 hybrid at 300 K (a) and 5 K (b) and thermal variation of zero field cooled (ZFC) and field cooled (FC) magnetization obtained under an applied field of 100 Oe (c) and 1000 Oe (d).

Table 3. Parameters of the Hysteresis Loops of CMK-3@m1 Hybrid Measured at a Temperature of 5 Ka sample

T (K)

Mmax+ (70 kOe) (emu/g)

Mmax‑ (70 kOe) (emu/g)

HC+ (Oe)

HC‑ (Oe)

MR+ (emu/g)

MR‑ (emu/g)

CMK-3@m1

5

2.12

−2.12

445

−554

0.64

−0.52

Mmax+ (70 kOe) is a maximum magnetization at 70 kOe, Mmax− (70 kOe) is a maximum magnetization at −70 kOe, HC+ is a positive coercivity, HC− is negative coercivity, MR+ is a positive remanent magnetization, and MR− is a negative remanent magnetization. a

3.3. Remediation Experiments for Cr 6+ Removal from Aqueous Solution. The efficiency of hybrids for removal of hexavalent chromium ions from aqueous solutions was investigated by conducting a series of kinetic experiments. Figure 9 shows the effectiveness of CMK-3@m1 in Cr(VI) removal at three initial concentrations (6, 2, and 0.6 mg/L) as a function of reaction time for two different pH values (5.5 and 3) at 25 °C. The analogous effectiveness of parent CMK-3 in Cr(VI) removal at concentration of pollutant of 6 mg/L, is also shown in the same figure for comparison purpose. It is evident that the reaction rate is faster at pH = 3 and the efficiency of sorbents for Cr(VI) removal is higher than that under less acidic conditions (pH = 5.5). Specifically, the efficiency of hybrid for chromium removal after 24 h reaction time at pH equals 5.5 was determined 89%, 66%, and 62% of the total chromium amount, as its concentration increases from 0.6 to 2 and finally 6 mg/L in the water solution. On the other hand, at pH = 3, the 100% of the total 0.6, 2, and 6 mg of Cr(VI) in the influent was already removed by the CMK-3@m1 at 1, 6, and 14 h, respectively. Furthermore, it is observed that, for given

constant hybrid dose (180 mg/L), the Cr(VI) removal rate decreases as the initial Cr(VI) concentration increases. The parent CMK-3 shows a significantly slower reaction rate and less efficiency at pH = 5.5 for Cr(VI) removal in comparison to that of hybrid, that was determined to be about 12% of Cr(VI) after 24 h of treatment. In contrast, its efficiency increases at pH = 3 so that more than 90% of the initial Cr(VI) was removed after the same elapsed time. The kinetics of the adsorption process shown in Figure 9a was simulated using a pseudo-second-order model expressed by the equation d([Cr(VI)]0 − [Cr(VI)]t ) = −k2([Cr(VI)]t )2 dt

(1)

where [Cr(VI)]0 and [Cr(VI)]t are the Cr(VI) concentrations in the solution after 0 and t hours of reaction and where k2 is the second-order rate constant (L·mg−1·h−1). By integrating eq 1 and using the boundary conditions of t = 0 to t and [Cr(VI)]t 3926

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

which indicates that the rate of chromium removal by sorbents fitted well the pseudo-second-order model under various initial concentrations of pollutant or different pH conditions. In addition the half time t1/2 can be calculated by the equation t1/2 = 1/(k2·[Cr(VI)]0 )

(3)

The kinetic data of Cr(VI) sorption have also been checked by first- and second-order models65,66 for the adsorbed material as function of time using nonlinear solutions of several kinetic models such as

q = qe(1 − exp−K1t )

q=

K2qe 2t 1 + K2qet

(5)

where Ki are rate constants (K1: min−1) or (K2: g/(mg min)) and qe is the equilibrium adsorption capacity (mg/g). The calculated R2 values indicate that better fit could be achieved with the second-order models (eq 5) in comparison with the first-order model (eq 4). However, the nonlinear regression analysis of the presented experimental data for pseudo-second-order models did not improve the fitting results significantly. The obtained R2 values with the nonlinear models were found to be satisfactory in some data sets; however, the same models provided low values of R2 ∼0.70, especially for the cases of small concentrations of Cr(VI) or when applied for the cases when the sorbent was the pristine CMK-3 matrix. Thus, it seems more appropriate to use the linear method to estimate the parameters involved in the kinetic eq 1. The rate of the of Cr(VI) removal from the water solution can be calculated from eq 2, since the rate r = k2([Cr(VI)]t)2 corresponds exactly to the reaction rate between the adsorbent and Cr(VI) species. Figure 9c shows the plot of reaction rate (r) as a function of concentration [Cr(VI)] for CMK-3@m1 hybrid (2 and 6 mg/L) and pristine CMK-3 (6 mg/L). It is observed that, at the beginning of the reaction and for the maximum [Cr(VI)] concentrations, the reaction rate of hexavalent chromium removal by hybrid material was not dependent on its initial concentration and was calculated as ∼14.5 mg·L−1·h−1, while in the case of the CMK-3, it was only 2.7 mg·L−1·h−1. In other words the hybrid material showed about 7 times higher reaction rate with [Cr(VI)] ions. However, the induced decrease of the chromium concentration by the adsorbents strongly affects the reaction rate (r); it is also decreased very rapidly. For concentrations less that 7 whereas that of γ-Fe2O3 oxide is generally higher (for magnetic oxides, PZC > 6.371). Thus, the hybrid surface’s active sites are in the form of both iron and carbon sites of the X−OH (X = Fe, C) type, all of which can be protonated to form X−OH2+ groups in the more acidic conditions as is the case during the batch experiments (i.e., pH = 5.5). In this case, the electrostatic attraction between anionic HCrO4− or Cr2O72− species and cationic surfaces may enhance the sorption of chromium on adsorbent, which could be generalized to the effect that both

Figure 9. (a) Effect of initial pH value (5.5 and 3) and Cr(VI) concentration (6, 2, and 0.6 mg/L) on the Cr(VI) removal efficiency by CMK-3 carbon and CMK-3@m1 hybrid. (b) Pseudo-second-order linear plot for the removal of Cr(VI) on CMK-3 carbon and CMK-3@ m1 hybrid. (c) Reaction rate as a function of Cr(VI) concentration of CMK-3 carbon and CMK-3@m1 hybrid.

= [Cr(VI)]0 to [Cr(VI)]t = [Cr(VI)]t, the linear equation is obtained 1 1 = k2t + [Cr(VI)]t [Cr(VI)]0

(4)

(2)

The plots of 1/Cr(VI) versus time produced linear plots (Figure 9b) with correlation coefficients (R2) higher than 0.935 3927

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

Table 4. Kinetics Constants for the Adsorption of Cr(VI) on the CMK-3@m1 Hybrid and CMK-3 Pristine Material samples pH = 3

CMK-3@m1

pH = 5.5

CMK-3 CMK-3-O@m4 CMK-3-O CMK-3@m1

CMK-3 CMK-3-O@m4 CMK-3-O

Cr6+ini (ppm)

R2

k2 × 10−3 (L·mg−1·h−1)

t1/2 (hours)

%E (24 h)

0.6 2 6 6 6 6 0.6 2 6 6 6 6

− 0.946 0.984 0.989 0.980 0.843 0.979 0.935 0.985 0.873 0.918 0.996

− 3359 434 82 185 6.36 562.07 35.94 10.98 0.806 5.63 3.28

− 0.14 0.4 2.1 0.9 27 0.3 13 16 215 31 53

100 100 100 92 96 51 89 66 62 12 48 31

has also been detected by XRF measurements. The fast kinetics of γ-Fe2O3 phase is due both to its nanometric size, which is realized as highly active iron particles (∼15 nm), as well as the existence of a high selectivity of γ-Fe2O3 to Cr(VI).71,72 The adsorption mechanism is mainly due to electrostatics attraction as the iron oxidation state being +3 in γ-Fe2O3 ensures that there is no redox taking place to reduce Cr(VI) to Cr(III). Thus, at low pH values, iron-oxide surfaces are protonated so that the net surface charge is positive, which enhances the adsorption of the negatively charged oxyanionic chromium species.71,73,74 On the other hand, the carbon surface of hybrids may act as appropriate substrates for reduction of Cr(VI) to Cr(III). Recently, that process was identified for hexavalent chromium reduction to Cr(III) by activated carbon material using synchrotron radiation-based X-ray emission spectroscopy.75 Therefore, the CMK-3 surface could also, acting as reducer, increase the adsorption capability of the hybrid material. It must be noted that simple pristine mesoporous carbon adsorption results show that the carbon surface on its own shows minor adsorption/reduction capability in comparison to that of the hybrid material. Along these lines, a deeper understanding of the adsorption mechanism is a subject of the consequent paper. Regeneration of a CMK/iron matrix is considered to be a possibility in further application of these materials, but it is possible upon such strong acid treatment to destroy the iron nanoparticles. In fact, XRF measurements reveal that ∼40 wt % of total iron oxide content of CMK-3 hybrids dissolves after 24 h treatment of these sorbents with acidic solutions (pH = 3) containing the HCrO4− or Cr2O72− species. On the other hand, NaBH4 could be used to regenerate the magnetic properties of nanoparticles forming zerovalent iron, which is also an important species for environmental applications, especially for Cr(VI) removal. This procedure would also reduce possible adsorbed Cr(VI) ions to Cr(III).

weak and electrostatic force adsorption is taking place on the hybrid surfaces. At lower pH values (pH = 3), the large number of H+ present produces more charged sites on the adsorbent surface resulting in a stronger electrostatic attraction of Cr(VI) anionic species and, thus, to an increase of adsorption efficiency of the hybrid. Moreover, CMK-3@m1 and CMK-3-O@m4 hybrids showed almost the same efficiency for Cr(VI) removal (Table 4), indicating that the size and, thus, the surface area of nanoparticles provide crucial dependencies toward the material’s sorption or reductive capacity, instead of the more natural quantity of total mass which is less important for the same properties. However, both hybrids exhibit lower surface areas in comparison to the pristine carbon (CMK-3) and both show better efficiency of Cr(VI) removal from the aqueous solutions as well as faster reaction kinetics. Gibbs free energy (ΔG0), an important thermodynamic parameter, was calculated for CMK-3@m1 hybrid at room temperature and pH = 5.5 using the equations KD = qe/Ce and ΔG0 = −RT ln KD. Ce is the equilibrium concentration (mg/ mL), qe is the amount of Cr(VI) adsorbed at equilibrium (mg/ g), KD is the distribution coefficient (mL/g), R is the gas constant (8.134 J/mol K), and T is the temperature in Kelvin. As calculated values of ΔG0 were negative, all confirm the spontaneous nature of the adsorption process. In particular, the Gibbs free energy was found to be −26.7 for 0.6 mg/L Cr(VI) initial concentration and −23.04 or −22.7 for 2 mg/L and 6 mg/L initial Cr(VI) concentrations, respectively. Similar reaction kinetics is being observed by the CMK-3O@m4. It is evident from the comparison of kinetic data concerning the CMK-3 pristine carbon and CMK-3@m1 hybrid (Figure 9) that in the case of hybrid material the overall mechanism is more complicated. In fact, it is obvious that the hybrids consisting of two sorption phases (mesoporous carbon and γ-Fe2O3 oxide nanoparticles) exhibit complicated behavior due to the concurrent different adsorption kinetics of each phase. The CMK-3 matrix slowly removes the Cr(VI) species toward immobilization, while γ-Fe2O3 nanoparticles probably react rapidly with chromium anion species. According to the WXRF results, the oxygen/carbon ratio results show that there is a dramatic increase (107%) in the oxygen content of CMK-3 after oxidation, while another significant increase (67%) occurs at the stage of iron oxide nanoparticle growth, occurring on the CMK-3 surface. It is very interesting that, once Cr(VI) solution treatment has taken place, the oxygen content increases again (28%) possibly pointing to the successful Cr adsorption. Similar effects take place upon pristine material treatment with Cr(VI) solution. Chromium atom content as adsorbed species

4. CONCLUSIONS Magnetic carbon hybrids composed of γ-Fe2O3 nanoparticles immobilized onto CMK-3 surfaces with variable iron oxide contents up to 27% have been prepared, characterized, and tested for Cr(VI) treatment. The synthesis method is based on the affinity of acetic acid vapors to react with the dispersed iron cations in the carbon matrix to first form iron acetate precursor species, which further produce the magnetic nanoparticles upon pyrolysis. This preparation method leads to iron oxide carbon hybrids, which combine stable ordered structures of interconnected carbon rods, high specific surface areas up to 1043 3928

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

m2/g, homogenously dispersed γ-Fe2O3 nanoparticles with uniform sizes (∼20 nm) and superparamagnetic properties (TB ≈ 50 K). Due to the specific surface properties of magnetic carbon hybrids and synergic effects of both counterparts, these exhibit extremely highly efficiency for removal of Cr(VI), considerably higher compared to that of pristine CMK-3 carbon. The adsorption kinetic data fitted well with a pseudosecond-order reaction model with linear method in order to adequately describe the kinetic uptake of Cr(VI) on both sorbents, i.e., CMK-3 and CMK-3@mx hybrids, in comparison to the applied nonlinear methods. The calculated values of Gibbs free energy were negative, confirming that the adsorption of Cr(VI) on CMK-3@mx hybrids is spontaneous and thermodynamically favorable.

■ ■

Bartonkova, H.; Bellesi, V.; Novak, P.; Petridis, D. Biomaterials 2009, 30, 2855−2863. (20) Dallas, P.; Tucek, J.; Jancik, D.; Kolar, M.; Panacek, A.; Zboril, R. Adv. Funct. Mater. 2010, 20, 2347−2354. (21) Beker, U.; Cumbal, L.; Duranoglu, D.; Kucuk, I.; Sengupta, A. K. Environ. Geochem. Health 2010, 32, 291−296. (22) Liu, W. T. J. Biosci. Bioeng. 2006, 102, 1−7. (23) Vatta, L. L.; Sanderson, R. D.; Koch, K. R. Pure Appl. Chem. 2006, 78, 1793−1801. (24) Hermanek, M.; Zboril, R.; Medrik, I.; Pechousek, J.; Gregor, C. J. Am. Chem. Soc. 2007, 129, 10929−10936. (25) Klimkova, S.; Cernik, M.; Lacinova, L.; Filip, J.; Jancik, D.; Zboril, R. Chemosphere 2011, 82, 1178−1184. (26) Huwe, H.; Fröba, M. Carbon 2007, 45, 304−314. (27) Minchev, C.; Huwe, H.; Tsoncheva, T.; Paneva, D.; Dimitrov, M.; Mitov, I.; Fröba, M. Microporous Mesoporous Mater. 2005, 81, 333−341. (28) Lee, J.; Jin, S.; Hwang, Y.; Park, J. G.; Park, H. M.; Hyeon, T. Carbon 2005, 43, 2536−2543. (29) Alam, S.; Anand, C.; Logudurai, R.; Balasubramanian, V. V.; Ariga, K.; Bose, A. C.; Mori, T.; Srinivasu, P.; Vinu, A. Microporous Mesoporous Mater. 2009, 121, 178−184. (30) Zhai, Y.; Dou, Y.; Liu, X.; Tu, B.; Zhao, D. J. Mater. Chem. 2009, 19, 3292−3300. (31) Dong, X.; Chen, H.; Zhao, W.; Li, X.; Shi, J. Chem. Mater. 2007, 19, 3484−3490. (32) Hu, L.; Dang, S.; Yang, X.; Dai, J. Microporous Mesoporous Mater. 2012, 147, 188−193. (33) Kim, B. C.; Lee, J.; Um, W.; Kim, J.; Joo, J.; Lee, J. H.; Kwak, J. H.; Kim, J. H.; Lee, C.; Lee, H.; Addleman, R. S.; Hyeon, T.; Gu, M. B.; Kim, J. J. Hazard. Mater. 2011, 192, 1140−1147. (34) Gu, Z.; Deng, B.; Yang, J. Microporous Mesoporous Mater. 2007, 102, 265−273. (35) Aggarwal, D.; Goyal, M.; Bansal, R. C. Carbon 1999, 37, 1989− 1997. (36) Jewur, S. S.; Kuriacose, J. C. Thermochim. Acta 1977, 19, 195− 200. (37) Pinheiro, E. A.; De Abreu Filho, P. P.; Galembeck, F.; Da Silva, E. C.; Vargas, H. Langmuir 1987, 3, 445−448. (38) Bourlinos, A. B.; Karakassides, M. A.; Simopoulos, A.; Petridis, D. Chem. Mater. 2000, 12, 2640−2645. (39) Kavanaugh, M. C. In Alternatives for Groundwater Cleanup; National Academic Press: Washington, DC, 1994; p 315. (40) Rai, D.; Eary, L. E.; Zachara, J. M. Sci. Total Environ. 1989, 86, 15−23. (41) Richard, F. C.; Bourg, A. C. M. Water Res. 1991, 25, 807−816. (42) Jun, S.; Sang Hoon, J.; Ryoo, R.; Kruk, M.; Jaroniec, M.; Liu, Z.; Ohsuna, T.; Terasaki, O. J. Am. Chem. Soc. 2000, 122, 10712−10713. (43) Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem. Soc. 1998, 120, 6024−6036. (44) Cao, J.; Zhang, W. X. J. Hazard. Mater. 2006, 132, 213−219. (45) Yang, Y.; Chun, Y.; Shang, G.; Huang, M. Langmuir 2004, 20, 6736−6741. (46) Dandekar, A.; Baker, R. T. K.; Vannice, M. A. Carbon 1998, 36, 1821−1831. (47) Bazula, P. A.; Lu, A. H.; Nitz, J. J.; Schüth, F. Microporous Mesoporous Mater. 2008, 108, 266−275. (48) Kathi, J.; Rhee, K. Y. J. Mater. Sci. 2008, 43, 33−37. (49) Kuznetsova, A.; Mawhinney, D. B.; Naumenko, V.; Yates, J. T. Jr; Liu, J.; Smalley, R. E. Chem. Phys. Lett. 2000, 321, 292−296. (50) Roggenbuck, J.; Waitz, T.; Tiemann, M. Microporous Mesoporous Mater. 2008, 113, 575−582. (51) Velasco-Santos, C.; Martínez-Hernández, A. L.; Lozada-Cassou, M.; Alvarez-Castillo, A.; Castaño, V. M. Nanotechnology 2002, 13, 495−498. (52) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordination Compounds, 3rd ed.; Wiley: New York, 1997; Part III (53) Kudin, K. N.; Ozbas, B.; Schniepp, H. C.; Prud’homme, R. K.; Aksay, I. A.; Car, R. Nano Lett. 2008, 8, 36−41.

AUTHOR INFORMATION

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work has been supported by the Operational Program Research and Development for Innovations − European Development Fund (CZ.1.05/2.1.00/03.0058), the project of the Grant Agency of the Czech Republic (P208/10/1742) and the project of the Academy of Sciences of the Czech Republic (KAN115600801). Helpful discussions with Assistant Professor I. B. Koutselas are also gratefully acknowledged.



REFERENCES

(1) Ryoo, R.; Joo, S. H.; Kruk, M.; Jaroniec, M. Adv. Mater. 2001, 13, 677−681. (2) Vinu, A.; Mori, T.; Ariga, K. Sci. Technol. Adv. Mater. 2006, 7, 753−771. (3) Shin, H. J.; Ryoo, R.; Kruk, M.; Jaroniec, M. Chem. Commun. 2001, 349−350. (4) Kruk, M.; Jaroniec, M.; Ryoo, R.; Joo, S. H. J. Phys. Chem. B 2000, 104, 7960−7968. (5) Li, L.; Zhu, Z. H.; Lu, G. Q.; Yan, Z. F.; Qiao, S. Z. Carbon 2007, 45, 11−20. (6) Sui, Q.; Huang, J.; Liu, Y.; Chang, X.; Ji, G.; Deng, S.; Xie, T.; Yu, G. J. Environ. Sci. 2011, 23, 177−182. (7) Wu, Z.; Webley, P. A.; Zhao, D. Langmuir 2010, 26, 10277− 10286. (8) Chandrasekar, G.; Son, W. J.; Ahn, W. S. J. Porous. Mater. 2009, 16, 545−551. (9) Lufrano, F.; Staiti, P. Energy Fuels 2010, 24, 3313−3320. (10) Zhang, J.; Kong, L. B.; Cai, J. J.; Luo, Y. C.; Kang, L. Electrochim. Acta 2010, 55, 8067−8073. (11) Liu, G.; Zheng, S.; Yin, D.; Xu, Z.; Fan, J.; Jiang, F. J. Colloid Interface Sci. 2006, 302, 47−53. (12) Vinu, A.; Hossain, K. Z.; Satish Kumar, G.; Ariga, K. Carbon 2006, 44, 530−536. (13) Kruis, F. E.; Fissan, H.; Peled, A. J. Aerosol Sci. 1998, 29, 511− 535. (14) Park, C. D.; Magana, D.; Stiegman, A. E. Chem. Mater. 2007, 19, 677−683. (15) Frey, N. A.; Peng, S.; Cheng, K.; Sun, S. Chem. Soc. Rev. 2009, 38, 2532−2542. (16) Fang, C.; Zhang, M. J. Mater. Chem. 2009, 19, 6258−6266. (17) Lu, A. H.; Salabas, E. L.; Schüth, F. Angew. Chem., Int. Ed. 2007, 46, 1222−1244. (18) Prucek, R.; Tuček, J.; Kilianová, M.; Panácě k, A.; Kvítek, L.; Filip, J.; Kolár,̌ M.; Tománková, K.; Zbořil, R. Biomaterials 2011, 32, 4704−4713. (19) Kluchova, K.; Zboril, R.; Tucek, J.; Pecova, M.; Zajoncova, L.; Safarik, I.; Mashlan, M.; Markova, I.; Jancik, D.; Sebela, M.; 3929

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930

Langmuir

Article

(54) Kim, T. W.; Park, I. S.; Ryoo, R. Angew. Chem., Int. Ed. 2003, 42, 4375−4379. (55) Fröba, M.; Köhn, R.; Bouffaud, G.; Richard, O.; Van Tendeloo, G. Chem. Mater. 1999, 11, 2858−2865. (56) Huwe, H.; Fröba, M. Microporous Mesoporous Mater. 2003, 60, 151−158. (57) Köhn, R.; Brieler, F.; Fröba, M. Ternary transition metal oxides within mesoporous MCM-48 silica phases: Synthesis and characterization. Stud. Surf. Sci. Catal. 2000, 129, 341−348. (58) Köhn, R.; Fröba, M. Catal. Today 2001, 68, 227−236. (59) Ennas, G.; Marongiu, G.; Musinu, A. J. Mater. Res. 1998, 14, 1570. (60) Greenwood, N. N.; Gibb, T. C. In Mössbauer Spectroscopy; Chapman & Hall: London, 1971. (61) Murad, E.; Johnston, J. H. Iron Oxides and Oxyhydroxydes, In Mössbauer Spectroscopy Applied to Inorganic Chemistry, Long, G. J., Eds.; Plenum Press: New York, 1987; Vol. 2. (62) Mørup, S.; Hansen, M. F. Superparamagnetic Particles, In Handbook of Magnetism and Advanced Magnetic Materials, Kronmüller, H., Parkin, S., Eds.; John Wiley & Sons, New York, 2007; Vol. 4 Novel Materials. (63) Jeong, J. R.; Lee, S. J.; Kim, J. D.; Shin, S. C. Phys. Stat. Solidi B 2004, 241, 1593−1596. (64) Dormann, J. L.; Fiorani, D.; Tronc, E. Magnetic relaxation in fine-particle systems. Adv. Chem. Phys. 1997, Vol. 98, 283−494. (65) Kumar, K. V. J. Hazard. Mater. 2006, 137, 1538−1544. (66) Ho, Y. S. J. Hazard. Mater. 2006, 136, 681−689. (67) Zink, S.; Schoenberg, R.; Staubwasser, M. Geochim. Cosmochim. Acta 2010, 74, 5729−5745. (68) Dakiky, M.; Khamis, M.; Manassra, A.; Mer’eb, M. Adv. Environ. Res. 2002, 6, 533−540. (69) Hsu, N. H.; Wang, S. L.; Lin, Y. C.; Sheng, G. D.; Lee, J. F. Environ. Sci. Technol. 2009, 43, 8801−8806. (70) Palmer, C. D.; Wittbrodt, P. R. Environ. Health Perspect. 1991, 92, 25−40. (71) Wang, P.; Lo, I. M. C. Water Res. 2009, 43, 3727−3734. (72) Goshu, I. V.; Tsarev, Y. V.; Kostrov, V. V. Russ. J. Appl. Chem. 2009, 82, 801−804. (73) Hu, J.; Chen, G.; Lo, I. M. C. Water Res. 2005, 39, 4528−4536. (74) Hu, J.; Lo, I. M. C.; Chen, G. Langmuir 2005, 21, 11173−11179. (75) Módenes, A. N.; Espinoza-Quinones, F. R.; Palácio, S. M.; Kroumov, A. D.; Stutz, G.; Tirao, G.; Camera, A. S. Chem. Eng. J. 2010, 162, 266−272.

3930

dx.doi.org/10.1021/la204006d | Langmuir 2012, 28, 3918−3930