Preparation of Biotic and Abiotic Iron Oxide Nanoparticles (IOnPs) and

sphere, and lithosphere (10). For this reason, their applica- tions have been studied in a number of diverse scientific disciplines, ranging from medi...
0 downloads 0 Views 393KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 4741-4747

Preparation of Biotic and Abiotic Iron Oxide Nanoparticles (IOnPs) and Their Properties and Applications in Heterogeneous Catalytic Oxidation HAERYONG JUNG,† HOSIK PARK,† JUN KIM,† JI-HOON LEE,† HOR-GIL HUR,† NOSANG V. MYUNG,‡ AND H E E C H U L C H O I * ,† Department of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), 1 Oryong-dong, Buk-gu 500-712, South Korea, and Department of Chemical and Environmental Engineering, University of CaliforniasRiverside, Riverside, California 92521

Iron oxide nanoparticles (IOnPs) as solid catalyst were prepared using a biotic method, i.e., biomineralization, and abiotic methods, i.e., thermal decomposition and electrochemical methods, for use as solid catalysts in the heterogeneous catalytic ozonation of para-Chlorobenzoic acid (pCBA). It was determined that characteristics of IOnPs, including particle size, morphology, surface area, electrokinetic mobility, basic group content, and chemical composition were significantly influenced by the preparation methods. TEM and FE-SEM analyses showed that the thermal decomposition method produced monodispersed and regularly spherical particles. The smallest iron oxide was also prepared by the thermal decomposition method, whereas the electrochemical method produced the largest iron oxide in terms of mean particle size. The specific surface area was found to be inversely proportional to the mean particle size. In catalytic ozonation at acidic pH levels, it was clearly observed that IOnPs enhanced the degradation of pCBA by the production of •OH radicals resulting from the catalytic decomposition of ozone. Additionally, functional groups and surface area were found to play an important role in the catalytic activity of IOnPs. To this extend, in a comparison of particle types, IOnPs prepared by the thermal decomposition method (IOTD) showed the greatest catalytic activity in terms of Rct value representing the ratio of hydroxyl radicals and ozone. This result may be due to the relatively higher surface area and basic group content of IOTD than other IOnPs.

Introduction Advanced oxidation processes (AOPs) have been developed to overcome the limitations of ozonation processes, such as the formation of byproducts and selective reactions of ozone, as AOPs are designed to enhance the production of hydroxyl * Corresponding author phone: +82-62-970-2441; fax: +82-62970-2434; e-mail: [email protected]. † Gwangju Institute of Science and Technology (GIST). ‡ University of CaliforniasRiverside. 10.1021/es0702768 CCC: $37.00 Published on Web 05/24/2007

 2007 American Chemical Society

radicals (•OH), known nonselective oxidants (1). However, another alternative is to use a metal catalyst for ozonation, as it can initiate the decomposition of aqueous ozone. Specifically, solid metal oxides are more practical in catalytic ozonation than ionized metals, due to the fact that solid metal oxides subdue bromate formation in the reaction with ozone and are less pH-sensitive than ionized metals (2, 3). Furthermore, it has been previously demonstrated that solid metal catalysts can be effectively utilized to improve ozonation efficiency in the removal of organic compounds (4-6). Iron oxides and hydroxides as solid metal catalysts are present in most regions of the different compartments of the global system, including the atmosphere, biosphere, hydrosphere, and lithosphere (10). For this reason, their applications have been studied in a number of diverse scientific disciplines, ranging from medicine, biology, and environmental chemistry to industrial chemical technology. For instance, in environmental chemistry, iron oxides and hydroxides have been used as both sorbent and catalyst in chemical oxidation due to advantages such as high stability. In catalytic ozonation, specifically, iron oxides and hydroxides have been commonly used (7, 8) based on advantages such as their abundance on Earth and their stable form (9). Beltran et al. (7) reported that alumina supported iron oxide (Fe2O3/ Al2O3) catalyst showed approximately 4 times higher catalytic activity than homogeneous Fe(III) catalyst in the catalytic ozonation of oxalic acid. Park et al. (8) also tested goethite (FeOOH) as a catalyst in the heterogeneous catalytic ozonation of para-Chlorobenzoic acid (pCBA). They concluded that the decay rates of ozone and pCBA were much higher than in the absence of goethite and were strongly pH dependent. Currently, iron oxide nanoparticles (IOnPs) are drawing a great deal of attention for the following reasons: (1) nanoparticles have a larger surface area/volume ratio than bulk materials for a given amount; and (2) the surface structure of IOnPs may differ from those of bulk materials. For these reasons, IOnPs have been applied in a number of research and industrial fields, including gas sensing applications (11), medical drug targeting materials in biomedicine (12), catalysts (13), magnetic storage media (14), and the adsorbent of environmental contaminants in soil and aqueous solutions (9). Moreover, IOnPs have been used to fabricate nanocomposites with carbon nanotubes and other metal oxides; the surface modification of a multiwalled carbon nanotube (MWCNT) with IOnPs enhanced the electrochemical activity of MWCNT for the electrochemical reduction of hydrogen peroxide (15). In other environmental applications, IOnPs have been applied in the adsorption of natural organic matter in aqueous solutions (9), and as a catalyst in heterogeneous catalytic oxidation (7). IOnPs have been synthesized by the electrochemical method, as well as the forced hydrolysis of sol-gel, thermal decomposition, and biomineralization methods. However, these preparation methods may result in the different surface properties of the resultant IOnPs. For instance, Nelson et al. (16) reported that preparation methods of iron and manganese (hydr)oxides significantly influenced the adsorption of lead; mainly caused by a difference in surface area. Thus, it is important to quantify the surface catalytic activity of IOnPs prepared by different methods. To the best of our knowledge, this is the first attempt to characterize the physicochemical properties of various IOnPs fabricated by different biotic and abiotic methods for application as catalysts in heterogeneous catalytic oxidation. VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4741

4742

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

NDe 6.3 aldrich IOAld_Fe(II,III)

d

Reference values from refs 10 and 32. c Reference value from ref 10.

aldrich IOAld_Fe(III)

a

Reference values from refs 25, 26, and 27.

b

electrochemical method IOEC

Surface area estimated by calculation. e Not detected.

45.1 48.6 1.1 60.0 (47.8)d 6.0 (4.2-7.0)b

6.6 ( 0.6

340.6

NDe 6.1 34.7 59.3 1.7 40.1 (40.5)d 16.4 ( 2.7 6.6 (6.6)c

378.9

0.7 3.7 40.5 55.2 1.4 31.5 (16.4)d 19.3 ( 2.4 8.4 (6.6)c

380.7

NDe 10.7 50.9 38.4 0.7 530.5 28.0 (31.6)d 10.0 ( 4.0 5.6 (4.2-7.0)b

Fe3O4 (magnetite) γ-Fe2O3 (maghemite) γ-Fe2O3 (maghemite) Fe3O4 (magnetite) IOBM

Cl

NDe 5.7

C O

37.4 56.9

Fe Fe/O

1.5 407.0 43.5 (43.8)d 7.2 ( 0.2 4.5 (6.1-9.5)a R-Fe2O3 (hematite)

basic group content (µeq g-1) BET surface area (m2 g-1) particle size (nm) pHiep mineral phase method or supplier name

TABLE 1. Preparation Methods and Characteristics of Iron Oxide Nanoparticles

Chemicals. All chemicals were obtained in high purity and were used as received. Iron(III) chloride (FeCl3‚6H2O) was supplied by Junsei (Japan). Ethanol and hexane were purchased from Duksan Chem. (South Korea). para-Chlorobenzoic acid (pCBA), oleic acid, and 1-octadecene were purchased from Aldrich (Milwaukee, Wisconsin), t-butanol was supplied by Fluka, and all solutions were prepared using deionized water (Ultrapure system, Barnsterd). Preparation of Iron Oxide Nanoparticles. IOnPs were prepared using a biotic method, i.e., biomineralization, and two abiotic methods, i.e., thermal decomposition and electrochemical methods. The preparation and supplier information are summarized in Table 1. Thermal Decomposition Method. The thermal decomposition method with respect to the preparation of IOnPs was reported in detail by Park et al. (18); however, the proposed method was slightly modified. In brief, iron(III) chloride and sodium oleate were first combined in a mixture solvent composed of ethanol, distilled water, and hexane. The solution was heated to 70 °C to prepare the iron-oleate complex. The iron-oleate complex was then dissolved in 200 g of 1-octandecene at room temperature. This mixture was subsequently heated up to 320 °C with a constant heating rate of 3.3 °C min-1, and the temperature maintained for 30 min. The solution was then cooled to room temperature, and excess ethanol was added to the solution, which yielded a black precipitate; the black precipitate was not stable in a polar solution such as water, because it was capped by an organic compound, oleic acid (18). To remove the capped oleic acid, the precipitates were redispersed in hexane and calcinated under an air-atmosphere at 550 °C for 5 h in a box-type electric furnace. After calcination, the powder turned red. Electrochemical Method. IOnPs were originally fabricated using the electrochemical method. In this paper, nanosized iron oxide was galvanostatically electrodeposited from an environmentally benign solution (i.e., 0.01 M FeCl3) at 20 °C. Briefly, 50 mL of 0.01 M FeCl3 solution controlled at pH 2 was added into a glass beaker. Then, a stainless steel sheet and steel rod to be used as the cathode and anode, respectively, were inserted into the reactor. The two electrodes were connected to a power supply, and an electric current was applied to the electrodes to fabricate IOnPs at 500 mA cm-2. The resulting nanoparticles were washed several times with nanopure water and then vacuum-dried. Biomineralization Method. Biogenic iron oxides can be formed by a diversity of organisms; however, there are two distinct methods for the synthesis of biogenic iron oxides: the biologically induced method, and the biologically controlled method. Note that biogenic iron oxide prepared by a biologically controlled method is a biomineral comprised of highly ordered crystals, whereas a biologically induced

thermal decomposition method and calcination in O2 biomineralization

normalized element composition (in weight percent)

Materials and Methods

IOTD

Based on the above, this paper focuses on (1) the synthesis and characterization of IOnPs for application as a heterogeneous catalyst in an oxidation reaction; and (2) the efficiency evaluation of IOnPs prepared by a variety of methods during heterogeneous catalytic ozonation. Three IOnP preparation methods of IOnPs were chosen in this study: the thermal decomposition method, the electrochemical method, and the biosynthesis method. For comparison, two commercially available IOnPs (Fe(II,III) and Fe(III)) were obtained from Aldrich (Milwaukee, Wisconsin). Here, the catalytic activity of IOnPs was estimated in the heterogeneous catalytic ozonation of para-chlorobenzoic acid (pCBA); pCBA was used as a target organic acid and probe compound for hydroxyl radicals due to its higher reactivity with hydroxyl radicals than ozone (k•OH/pCBA ) 5.2 × 109 M-1 s-1, kO3/pCBA ) 0.15 M-1 s-1) (17).

method produces extracellularly produced biomineral that does not crystallize under strict genetic controls (19). The amount of biogenic iron oxide prepared by the biologically controlled method is considerably less than the biologically induced method because it is synthesized within the microorganisms. For this study, we used a biologically induced method utilizing iron-reducing bacterium. Biogenic iron oxide was produced by a dissimilatory iron-reducing bacterium, Shewanella sp. The bacterium was incubated in an anaerobic condition for 1 month at 30 °C, in the presence of β-FeOOH as a precursor; details of which are described elsewhere (20). Characterization of IOnPs. The synthesized nanoparticles were characterized by transmission electron microscopy (TEM), X-ray diffraction (XRD), field emission scanning electron microscopy-energy dispersive X-ray (FE-SEM-EDX), Fourier transform infrared spectroscopy (FT-IR), and BET surface area measurements. TEM analyses were performed using a JEOL TEM 2000FXII operated at 200 kV. First, iron oxide samples previously distributed in ethanol were mounted on cupper grids; these cupper grids were kept in an oven at 100 °C prior to insertion into the microscope. Next, X-ray diffraction analyses were conducted to characterize the mineral phase of the iron oxide samples, which were subsequently analyzed in ambient air with Cu KR using a Rigaku RINT2000 wide-angle goniometer operated at 40 kV and 40 mA. Continuous scans from 10 to 80° 2θ were then collected. The Brunauer-Emmett-Teller (BET) surface areas were determined from N2 physisorption with an ASAP 2020, using BJH (Barrett-Joyner-Halenda) and multi-point BET methods. Note that before each measurement, samples were pretreated at 300 °C for 8 h. EDX analysis was carried out to estimate the normalized element composition of IOnPs using FE-SEM (S-4700, Hitachi, Japan) with EDX (Horiba, Japan). The EDX spectrometer was equipped with a lithium-drifted silicon detector of ultra thin window which allows X-ray detection from the elements with atomic number higher than that of berrylium (Z > 4). The electrokinetic mobility of the iron oxide samples in aqueous solutions was measured using an electrophoretic light scattering spectrophotometer (ELS-8000, Otsuka, Japan). Here, 5 mM of diluted iron oxide nanoparticles in an NaCl electrolyte solution were used at room temperature. Volumetric titrations were performed for aqueous solutions of IOnPs using a 702 SM Titrino (Metrohm, Swiss). The titrations were done to estimate the functional group contents of nanosized iron oxide at a constant temperature of 25 °C with 50 mL of diluted 5 mM iron oxide suspensions in deionized water. Catalytic Ozonation Experiments. Batch experiments were carried out with a 1 L reactor, as shown in the schematics of experimental setup (See Figure S1, Supporting Information). Briefly, a 1.0 × 10-2 M IOnP solution was prepared by adding an exact amount of IOnPs to deionized water. Then, the desired volume of pCBA and IOnPs solution was added to the reactor. Finally, the solution was continuously stirred with a Teflon stirrer at 150 rpm to disperse IOnPs, and the desired concentration of gaseous ozone was injected into the reactor. The use of a phosphate buffer was avoided in order to control the pH level, as the phosphate buffer could hinder the catalytic reaction on the surface of the catalyst. For this reason, either HCl or NaOH solution was used to maintain the pH level during the reaction. As shown in Figure S1, gaseous ozone was generated from pure oxygen by an ozone generator (PCE-WEDECO, GL-1, U.S.). After a designated time interval, 3 mL of the solution in the reactor was sampled to determine the concentration of pCBA. For pCBA analysis, an aliquot of 0.1 N Na2S2O3 was subsequently added to the sample to quench the aqueous

ozone remaining in the reaction solution. Triplicate experiments were conducted at 25 ( 1 °C for verification of all results. Error bars in the figures represent the confidential interval (R ) 0.05) based on triplicate experiments. pCBA was measured by using high performance liquid chromatography (HPLC, WATERS, U.S.) with an autosampler and a UV absorbance detector (Younglin, UV 730D, South Korea). Details are mentioned elsewhere (21).

Results and Discussion Structure, Morphology, and Composition. XRD, FE-SEMEDX, TEM, and BET measurements were performed to characterize the size, particle shape, surface area, and chemical composition of IOnPs. The results are summarized in Table 1 for which discussion will be made later part of this section. Particle Size and Morphology. XRD analysis revealed that IOnPs synthesized by the thermal decomposition method (IOTD) belong to a hematite phase (R-Fe2O3) (See Figure S2, Supporting Information). In this method, the black precipitates formed before calcination were magnetite (Fe3O4) capped with oleic acid. However, the phase changed to hematite during calcination under the air-atmosphere:

4Fe3O4 + O2 T 6Fe2O3

(1)

The biogenic iron oxide (IOBM) and Fe(II,III) from Aldrich (IOAld-Fe(II,III)) were magnetite phase (Fe3O4), a ferromagnetic mineral containing both Fe(II) and Fe(III). Though biologically induced iron oxide is typically poorly crystalline and impure (18), IOBM prepared for this study was well crystallized. Here, Fe(III) from Aldrich (IOAld-Fe(III))and iron oxide prepared by the electrochemical method (IOEC) displayed a maghemite (γ-Fe2O3) phase, which is isostructural with magnetite, but having cation deficient sites. TEM images were employed to examine the average size and morphology of IOnPs as shown in Figure 1, and the size histograms of IOnPs are shown in Figure S3 (See Supporting Information). Note that the thermal decomposition method only produced monodispersed and regularly spherical particles, as shown in Figure 1. The smallest particles were also prepared in the method summarized in Table 1. It was determined that nucleation and growth processes take place at different temperatures in the thermal decomposition method, suggesting that it is relatively easy to control the size of the nanoparticles. Furthermore, the oleic acid used in the synthesis process was found to play a significant role in preventing the aggregation of the nanoparticles. In terms of sizes, IOBM was primarily composed of irregular nanosized particles, average size 10.0 ( 4.0 nm in diameter, with particles aggregated into larger sizes. The coefficients of variance, i.e., the standard deviation relative to the mean, of IOTD, IOBM, IOEC, IOAld_Fe(III), and IOAld_Fe(II,III) were 0.03, 0.40, 0.12, 0.16, and 0.09, respectively, indicating that the size of IOBM varies more than those of other IOnPs. For example, the size variation of IOBM is approximately 13.3 times greater than for IOTD. This large size variance in size is a general characteristic of extracellularly produced biominerals (20). In this study, IOEC showed the largest average size and a relatively large size variation, but the average particle size and size variation of IOEC can be decreased by controlling the current density and electrolyte concentration, because particle size in the electrochemical method is inversely proportional to current density and electrolyte concentration. However, one significant problem is that increasing the current density and electrolyte concentration is not a costeffective fabrication method, as the current efficiency for IOnP fabrication significantly decreases with an increase in the current density and electrolyte concentration. In IOEC, VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4743

FIGURE 1. TEM images of iron oxide nanoparticles. the aggregation of IOnPs was observed in a chain-like structure, as shown in Figure 1. The surface morphology of IOnPs would also influence the catalytic activity in the heterogeneous catalytic ozonation. However, Figure 1 does not show significant difference in the surface morphology of IOnPs, and the surface of IOnPs is rather smooth with the exception of IOBM. The relatively irregular surface of IOBM does not enhance the catalytic activity in this study, based on the fact that IOBM showed the lowest degradation rate of pCBA in the heterogeneous catalytic ozonation. The catalytic activity of IOnPs will be discussed in detail later. Surface Area. Surface area is one of the essential characteristics of a solid catalyst, as catalytic activity is generally proportional to the surface area of the solid catalyst in heterogeneous catalytic ozonation. In this study, the BET method was used to estimate the specific surface area of IOnPs based on the assumption that IOnPs do not have an internal surface area. The specific surface area was also calculated under the further assumption that the particles have spherical geometry, as presented in Table 1. Here, the densities of hematite, magnetite, and maghemite were assumed to be 5.26, 5.18, and 4.87 g cm-3, respectively (10). 4744

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

The calculated specific surface area was then compared with the BET-determined surface area. It was determined that the calculated surface area matches reasonably well with the specific surface area measured by the BET method for all IOnPs. Specifically, for IOTD the specific surface areas by both the BET and calculated methods were estimated to be 43.5 and 43.8 m2 g-1, respectively. This good agreement may be due to the homogeneous distribution of particle size and the regular spherical geometry of IOTD particles. For IOBM, the specific surface areas estimated by the calculation and BET methods were 31.6 and 28.0 m2 g-1, respectively. The difference is not overly significant, although it showed the greatest value in the coefficient of variance. This result confirms that the assumption employed in the estimation of the BET method is acceptable. Chemical Composition. FE-SEM-EDX analyses identified the normalized elemental composition of IOnPs. EDX data reveals that IOnPs are mostly composed of Fe, O, and C, with the exception of IOEC, which contains Cl, originating from the ferric chloride used in the preparation of the sample, as listed in Table 1. For IOTD, IOEC, IOAld_Fe(III), the Fe/O ratios were 1.5, 1.4, and 1.7, respectively, based on weight percent. These values are significantly less than the ratio of Fe2O3

(approximately 2.32) and Fe3O4 (approximately 3.65). This relatively low Fe fraction and high O fraction may be due to the oxygenated functional groups and/or some carbon fractions formed during the manufacturing process as IOnPs contain C-O and aldehyde peaks in FT-IR spectra (See Figure S4, Supporting Information). Furthermore, IOBM also exhibited significantly higher O and C fractions, and a lower Fe fraction than Fe3O4. The higher O fraction of IOBM may be explained by the oxygenated functional groups on the particle surface, such as O-O from peroxidase and O-H from the carboxylic acid functional group. This result is also confirmed by FT-IR spectra; the highest C fraction of IOBM seems to originate from the organic residuals of microorganisms. Surface Functional Properties. The catalytic efficiency of solid catalysts greatly depends on the catalyst and its surface properties. To this extent, Kasprzk-Hordern et al. (22) have suggested that important properties of solid catalysts include surface area, mechanical strength, and the presence of active sites, such as acid and base sites, but concluded that the main parameters for the catalytic properties of metal oxides are acidity and basicity. SanchezPolo et al. (23) further reported that the surface contents of the basic groups of solid catalysts are essential in enhancing the transformation of ozone to hydroxyl radicals in heterogeneous catalytic ozonation due to the electrophilic characteristics of aqueous ozone. Therefore, the surface electric and functional properties of IOnPs are considered essential factors for determining the effectiveness of solid catalysts in heterogeneous catalytic ozonation. Surface Chemistry. In aqueous solutions, the surface of IOnPs is covered with hydroxyl groups as follows:

tFe-OH+2 T tFe-OH + H+ pK1

(2)

tFe-OH T tFe-O- + H+ pK2

(3)

Note here that the surface chemistry of the solid catalyst is strongly influenced by the pH levels of the solution in heterogeneous catalytic ozonation. As the pH level of solution goes over pK2 value, the surface of IOnPs becomes negatively charged. Ozone decomposition rates due to catalytic surface reaction are also highly influenced by solution pH levels. This is attributed to the different reactivity of aqueous ozone toward the three potential surface charges, i.e., positive, neutral, and negative charges, and the fact that the surface charge of a solid catalyst as shown in eqs 2 and 3, is pHdependent (8); it has been shown that a negatively charged solid catalyst has much higher reactivity toward ozone than either neutral or positively charged catalysts. This is likely due to the electrophilic characteristics of ozone, which has a high affinity for molecular sites with a strong electronic density. Therefore, the solid catalysts that maintain a negative surface charge potential in neutral pH levels could be a viable candidate for heterogeneous catalytic ozonation. To examine the surface charge properties of IOnPs, isoelectric points (pHiep) were determined in an NaCl electrolyte solution, as summarized in Table 1 (also see Figure S4, Supporting Information). Note that the preparation methods of IOnPs strongly influenced pHiep. For example, pHiep of all IOnPs, including the commercial ones, ranged from 4.5 to 8.4. These values are comparable with previous reports with the exception of IOTD, which showed the lowest pHiep, i.e., 4.5, an extraordinary low value when compared to previous reports which present the pHiep of IOTD in the range of 6.1-9.5 (24-26). This considerable difference seems to be caused by differences in experimental conditions, particle size, and preparation methods. Previous reports generally used submicrometer-sized hematite prepared by

FIGURE 2. Effect of iron oxide concentrations on pCBA degradation (IOTD). the forced hydrolysis of an acidic iron(III) solution and the aging of a concentrated iron hydroxide gel, although the mean size of hematite prepared in this study is less than 10 nm. Thus, this result implies that reducing the particle size of hematite influences the electrokinetic properties of particles. In fact, IOTD is potentially the best candidate for a solid catalyst in nanosized iron oxides with respect to pHiep value. Basic Group Content. The basic group contents of IOnPs range from 380 to 530 µeq g-1 (see Table 1). This is comparable with those of commercial activated carbons; Sanchez-Polo et al. (23) previously reported that two commercially available activated carbons used in ozonation experiments had basic group contents ranging between 253 and 570 µeq g-1. They pointed out that the activated carbon successfully applied as a catalyst in heterogeneous catalytic ozonation and catalytic activity was proportional to the basic group content. As such, this implies that IOnPs used in this study could be good candidates for a solid catalyst in heterogeneous catalytic ozonation as the basic group content of IOTD was 407.0 µeq g-1, whereas the basic group content of IOBM was determined to be 530.0 µeq g-1, the highest value for IOnPs used in this study. This may be due to residual biomaterials on the surface of IOBM, as even though IOBM was carefully washed several times with deionized water before use in the experiments to remove residual biomaterials, some fractions of the biomaterials were not removed. This existence of residual biomaterials was confirmed with FT-IR analysis (See Figure S4, Supporting Information). In our study, IOBM displayed much larger functional groups than the other iron oxides. Degradation of pCBA. Catalytic ozonation was carried out at acidic pH levels (approximately 2.5) in order to evaluate the surface-catalyzed reaction of ozone in the presence of IOnPs. The initial pCBA concentration was controlled at 1.28 × 10-2 mM. Figure 2 presents the temporal variations of pCBA concentrations at various concentrations of IOTD. It was found that the removal rate of pCBA increased with an increase in the concentration of IOTD. Other IOnPs also showed the same removal pattern of pCBA with respect to the concentration of IOnPs (data not shown here), suggesting that the production of hydroxyl radicals was accelerated by the catalytic reaction of ozone on the surface of IOnPs. As such, the removal of pCBA can be divided into two-phases: an initial rapid removal phase of pCBA (Phase I), followed by a rather slow decomposition phase (Phase II). Many researchers have reported this pattern of two-stage kinetics in the decomposition of phenanthrene (27), TCE (28), and pCBA (23, 29) in heterogeneous catalytic ozonation. The concentration of ozone off-gas was monitored during the reaction to examine the ozone demand of IOnPs, as VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4745

experiments without catalysts by the ozone off-gas concentrations in experiments containing each catalyst. Note that in an attempt to estimate Rct values from the slope, each curve in the figure was simply divided into two subsections with linear regions, as shown in Figure 4. Rct values of IOnPs are summarized in Table S1 (see the Supporting Information). For IOnPs, the Rct values during the first period (1 min) were determined to be much greater than those determined for the second period. For example, the Rct value was estimated to be 4.3 × 10-6 during the first period (1 min), and then it changed to 1.4 × 10-7 during the second period for IOBM, under our experimental conditions. This change of Rct seems to be caused by the initial rapid decrease of pCBA, as mentioned above. This pattern of two-stage kinetics has already been reported in the heterogeneous catalytic ozonation of pCBA using iron oxide and hydroxide (8) and nanosized ZnO (21). FIGURE 3. Change of gaseous ozone concentration during reaction.

FIGURE 4. Effect of catalysts on change of Rct (conc. of gaseous ozone, pCBA, and iron oxide ) 2.08 × 10-1 mM, 1.28 × 10-2 mM, and 56 mg L-1, respectively). presented in Figure 3. As can be seen, the concentration of gaseous ozone in the control experiment sharply increased within 1 min, and reached a steady state within 7 min of initiating the reaction. However, the saturation pattern of gaseous ozone in the presence of IOnPs levels 65-85% of the control experiment at 10 min after the reaction started, with a gradual increase from that time onward. It took a much longer time for the ozone concentration to reach a steady state in the presence of IOnPs. This result confirms that the surface-catalyzed ozone decomposition continually occurs during the reaction. Note that IOBM showed the highest ozone demand among IOnPs used in these experiments; although IOBM exhibited the lowest catalytic activity in heterogeneous catalytic ozonation (see Figure 4). This may be due to the scavenger effects of free radicals of organic matter on the particle surface. Other IOnPs displayed a similar ozone demand. Elovitz and von Gunten (30) proposed the Rct concept representing the ratio of the exposures of •OH radicals and ozone as follows:

(

ln

)

[pCBA] ) - k•OH/pCBARct [pCBA]o

∫[O ]dt 3

(4)

Rct values were evaluated in the catalytic ozonation in the presence of IOnPs. In the calculation of Rct values, the amount the aqueous ozone varied during the reaction was estimated by subtracting the ozone off-gas concentrations in control 4746

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 13, 2007

In a comparison of types of IOnPs, IOTD showed the greatest catalytic activity in IOnPs, based on Rct value. This result may be due to the relatively high surface area and basic group content of IOTD, as it is well-known that the catalytic effect in heterogeneous catalytic ozonation is proportional to the surface area and basic group contents of the catalyst. However, for IOBM, although it exhibited the greatest basic group content, it revealed the lowest Rct value, as the catalytic activity of IOBM was found to be lower than that of IOAld_Fe(II,III), which exhibited a higher surface area than IOBM. This result may be due to the fact that IOBM has the lowest surface area and highest organic residual content, especially as organic matter has been known to be a powerful scavenger of hydroxyl radicals during ozonation. Previous studies have reported that organic matter is the highest naturally occurring ozone and hydroxyl radical consuming matter in heterogeneous catalytic ozonation (32, 33). It was clearly observed that the surface properties of IOnPs were strongly influenced by the preparation methods, and that the catalytic activity of IOnPs was affected by the surface properties of IOnPs in heterogeneous catalytic ozonation. Moreover, the thermal decomposition method was found to produce highly effective monodispersed IOnPs in heterogeneous catalytic ozonation. This was primarily due to properties of IOTD, such as smaller particle size, higher surface area, lower pHiep, and higher basic group content, as compared to other IOnPs. However, the mineral phase effect of IOnPs on catalytic activity was not considered in this study, although three different mineral phases were considered. This was because it is very difficult to fabricate IOnPs with identical physicochemical properties in each of the different mineral phases. Reaction Mechanism of Ozone and pCBA. In the heterogeneous catalytic ozonation, the reaction mechanism is still unveiled because the reaction pathway is so complex. However, Park et al. (8) and Jung and Choi (21) proposed the possible reaction mechanisms for the degradation of ozone and pCBA in the heterogeneous catalytic ozonation of pCBA based on the experimental data and on a review of the literature. In this study, the possible mechanisms are summarized in Table S2 (see the Supporting Information). The reactions are initiated by the formation of a precursor surface complex (≡FeOH(O3)s or ≡FeO(O3)s-) of ozone on the surface of IOnPs. Then, the reactions involve a series of chain reactions from reaction (I-2) to (I-6), and hydroxyl radicals are formed. The main limiting step of pCBA decomposition involves the adsorption of pCBA on the surface of IOnPs. As shown in reaction (II-1), acidity plays a significant role in the adsorption of pCBA, because of outer sphere complexation between the carboxylate and hydroxyl groups (8). The adsorbed pCBA reacts with the hydroxyl radicals generated by the ozone surface reaction.

Acknowledgments This work was supported by the Korea Science and EngineeringFoundation(KOSEF)grant(no.M1050000012806J000012810) through the National Research Laboratory Program by the Korea government (MOST).

Supporting Information Available Rct values of different IOnPs are listed in Table S1. Reaction mechanisms of ozone and pCBA are summarized in Table S2. Experimental setup and other properties of IOnPs such as XRD results, FT-IR spectra, and isoelectric points (pHiep) are shown in Figures S1-S4, respectively. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Ozone in Water Treatment: Application and Engineering; Langlais, B., Reckhow, D. A., Brink, D. R., Eds.; Lewis Publishers: Chelsea, MI, 1991. (2) Lin, S.-S.; Gurol, M. D. Catalytic ozonation: a promising advanced oxidation technology for water treatment. Environ. Sci. Technol. 1998, 32, 1417-1423. (3) Dionysiou, D. D.; Suidan, M. T. Effect of ionic strength and hydrogen peroxide on the photocatalytic degradation of 4-chlorophenzoic acid in water. Appl. Catal., B 2000, 26, 153-171. (4) Andreozzi, R.; Insola, A.; Caprio, V.; Marotta, R.; Tufano, V. The use of manganese dioxide as a heterogeneous catalyst for oxalic acid ozonation in aqueous solution. Appl. Catal., A 1996, 138, 75-81. (5) Legube, B.; Leitner, N. K. V. Catalytic ozonation: a promising advanced oxidation technology for water treatment. Catal. Today 1999, 53, 61-72. (6) Beltran, F. J.; Rivas, F. J.; Montero-de-Espinosa, R. Ozoneenhanced oxidation of oxalic acid in water with cobalt catalysts. 2. heterogeneous catalytic ozonation. Ind. Eng. Chem. Res. 2003, 42, 3218-3224. (7) Beltran, F. J.; Rivas, F. J.; Montero-de-Espinosa, R. Iron type catalysts for the ozonation of oxalic acid in water. Water Res. 2005, 39, 3553-3564. (8) Park, J.-S.; Choi, H.; Cho, J. Kinetic decomposition of ozone and para-chlorobenzoic acid during catalytic ozonation. Water Res. 2004, 38, 2285-2292. (9) Waychunas, G. A; Kim, C. S.; Banfiel, J. F. Nanoparticulate iron oxide minerals in soils and sediments: unique properties and contaminant scavenging mechanisms. J. Nanopart. Res. 2005, 7, 409-433. (10) Cornell, R. M.; Schwertmann, U. The Iron Oxides, Structure, Properties, Reactions, Occurrences and Uses, 2nd ed; Wiley-VCH, Verlag GmbH & Co: Weinheim, Germany, 2003. (11) Rumyantseva, M.; Kovalenko, V.; Gaskov, A.; Makshina, E.; Yuschenko, V.; Ivanova, I.; Ponzoni, A.; Faglia, G.; Comini, E. Nanocomposites SnO2/Fe2O3: Sensor and catalytic properties. Sens. Actuators, B 2005, 118, 208-214. (12) Walczak, P.; Ruiz-Cabello, J.; Kedziorek, D. A.; Gilad A. A.; Lin, S.; Barnett, B.; Qin L.; Levitsky, H.; Bulte, J. W. M. Magnetoelectroporation: improved labeling of netural stem cells and leukocytes for celluar magnetic resonance imaging using a single FDA-approved agent. Nanomedicine 2006, 2, 89-94. (13) Zaki, T. Catalytic dehydration of ethanol using transition metal oxide catalysts. J. Colloid Interface Sci. 2005, 284, 606-613. (14) Sun, S.; Murray, C.; Weller, D.; Folks, L., Moser, A. Monodisperse FePt nanoparticles and ferromagnetic FePt nanocrystal supperlattices. Science 2000, 287, 1989-1992. (15) Sljukic, B.; Banks, C. E.; Compton, R. G. Iron oxide particles are the active sites for hydrogen peroxide sensing at multiwalled carbon nanotube modified electrodes. Nano Lett. 2006, 6, 15561558.

(16) Nelson, Y. M.; Lion, L. W.; Shuler, M. L.; Ghiorse, W. C. Effect of oxide formation mechanism on lead adsorption by biogenic manganese (hydro) oxides, iron (hydro) oxides, and their mixtures. Environ. Sci. Technol. 2002, 36, 421-425. (17) von Gunten, U.; Driedger, A.; Gallard, H.; Salhi, E. By-products formation during drinkwater disinfection: a tool to assess disinfection efficiency. Water Res. 2001, 35, 2095-2099. (18) Park, J.; An, K.; Hwang, Y.; Park, J.-G.; Noh, H.-J.; Kim, J.-Y.; Park, J.-H.; Hwang, N.-M.; Hyeon, T. Ultra-large-scale synthesis of monodisperse nanocrystals. Nature Mat. 2004, 3, 891895. (19) Weiss, F. R.; Kim, S. S.; Kirschvink, J. L.; Kopp, R. E.; Sankaran, M.; Kobayashi, A.; Komeili, A. Ferromagnetic resonance and low-temperature magnetic tests for biogenic magnetite. Earth Planet. Sci. Lett. 2004, 224, 73-89. (20) Lee, J.-H.; Roh, Y.; Kim, K.-W.; Hur, H.-G. Organic aciddependent iron mineral formation by a newly isolated ironreducing bacterium, Shewanella sp. HN-41, Geomicrobiol. J. 2006, in revision. (21) Jung, H.; Choi, H. Catalytic decomposition of ozone and paraChlorobenzoic acid (pCBA) in the presence of nanosized ZnO, Appl. Catal., B 2006, 66, 288-294. (22) Ernst, M.; Lurot, F.; Schrotter, J.-C. Catalytic ozonation of refractory organic model compounds in aqueous solution by aluminum. Appl. Catal., B 2004, 47, 15-25. (23) Sanchez-Polo, M.; von Gunten, U.; Rivera-Utrilla, J. Efficiency of activated carbon to transform ozone into ‚OH radicals: influence of operational parameters. Water Res. 2005, 39, 31893198. (24) Kallay, N.; Dojnovic´, Z.; CÄ op, A. Surface potential at the hematitewater interface. J. Colloid Interface Sci. 2005, 286, 610-614. (25) Schudel, M.; Behrens, S. H.; Holthoff, H.; Kretzschmar, R.; Borkovec, M. Absolute aggregation rate constants of hematite particles in aqueous suspensions: a comparison of two different surface morphologies. J. Colloid Interface Sci. 1997, 196, 241253. (26) Hsi, C. D.; Langmuir, D. Adsorption of uranly onto ferric oxyhydroxides: application of the surface complexation sitebinding model. Geochim. Cosmochim. Acta 1985, 49, 19311941. (27) Cooper, C.; Burch, R. An investigation of catalytic ozonation for the oxidation of halocarbons in drinking water preparation, Water Res. 1999, 33, 3695-3700. (28) Hwang, W.-J.; Fang, G.-C.; Wang, C.-C. A nanometer-ZnO catalyst to enhance the ozonation of 2,4,6-trichlorophenol in water. Colloids Surf. A 2005, 260, 45-51. (29) Lim, H.-L.; Choi, H.; Hwang, T.-M.; Kang, J.-W. Characterization of ozone decomposition in a soil slurry: kinetics and mechanism, Water Res. 2004, 36, 219-229. (30) Elovitz, M. S.; von Gunten, U. Hydroxyl radical/ozone ratios during ozonation processes. I. the Rct concept, Ozone: Sci. Eng., 1999, 21, 239-260. (31) Erdemoglu, M.; Sarikaya, M. Effects of heavy metals and oxlate on the zeta potential of magnetite. J. Colloid Interface Sci. 2006, 300, 795-804. (32) Jung, H; Kim, J.; Choi, H. Reaction kinetics of ozone in variably saturated porous media. ASCE J. Environ. Eng. 2004, 130, 432441. (33) Kim, J.; Choi, H. Modeling in situ ozonation for the remediation of nonvolatile PAH-contaminated soils. J. Contam. Hydrol. 2002, 55, 261-229.

Received for review February 4, 2007. Revised manuscript received April 4, 2007. Accepted April 12, 2007. ES0702768

VOL. 41, NO. 13, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4747