Pd Nanoparticle Immobilization in Microfiltration Membrane Pores

2348

Ind. Eng. Chem. Res. 2007, 46, 2348-2359

Fe/Pd Nanoparticle Immobilization in Microfiltration Membrane Pores: Synthesis, Characterization, and Application in the Dechlorination of Polychlorinated Biphenyls Jian Xu and Dibakar Bhattacharyya* Department of Chemical and Materials Engineering, UniVersity of Kentucky, Lexington, Kentucky 40506-0046

Nanosized Fe/Pd bimetallic particles are an important category of materials in the area of reductive degradation of chlorinated organics. In this work, we prepared Fe/Pd nanoparticles in three steps: polymerization of acrylic acid (AA) in poly(vinylidene fluoride) (PVDF) microfiltration membrane pores, subsequent ion exchange of Fe2+, and chemical reduction (by borohydride) of ferrous ions bound to the carboxylic acid groups. Fe/Pd bimetallic nanoparticles were formed by the partial reduction of Pd2+ with Fe0 nanoparticles. The functionalized membrane and the nanoparticles were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The membrane-supported nanoparticles exhibited high reactivity in the dechlorination of 2,2′-dichlorobiphenyl (DiCB) used as a model compound. The dechlorination mechanism and the role of water were probed by conducting the reaction in pure ethanol solution. Bulk Fe/Pd particles were also prepared to investigate the effect of particle size on the catalytic activity. The effect of Pd content on the catalytic activity was studied to understand and quantify the role of Pd in the bimetallic nanoparticle system. The high catalytic activity of Pd was confirmed by the low activation energy compared to those other catalytic systems. Introduction Nanosized metal particles have become an important class of materials because of their unique physical and chemical properties unlike those of the bulk metals. Substantial studies of metal nanoparticle synthesis have been reported in the fields of catalysis1 and optical,2 electronic,3 magnetic,4 and biological5,6 devices. In these cases, to avoid agglomeration and aggregation, the nanoparticles were usually stabilized by polymers or ligands in the solution phase or immobilized on solid supports.7-9 Much attention has been given to the preparation of metal nanoparticles embedded in thin and dense films or gels by a stepwise approach involving ion exchange and reduction.10-14 In this method, ionexchange ligands created in the thin films can bind metal cations from aqueous solution. Postreduction or precipitation produces nanoparticles from bound metal cation precursors. The advantage of this process is that the nanostructure properties can be controlled during nanoparticle synthesis by utilizing these ionexchange ligands. The nanostructure properties include the particle size and distribution, particle concentration, and interparticle spacing. The amount of metal cations loaded is controlled by the amount of ligand sites and ion-exchange conditions such as the pH and competitive ions.11 The distance between bound cations, which determines the final particle size, is also controlled by the spacing between ion-exchange ligands.15 For example, Wang and co-workers reported that silver nanoparticles with various sizes and concentrations were produced in the polyelectrolyte multilayer film.15 However, few studies have been conducted on the immobilization of nanoparticles in microporous membranes modified with ion-exchange ligands. Modification of microporous membranes with ion-exchange groups has received extensive attention for applications in heavy-metal capture,16 pervaporation,17 reverse osmosis,18 nanofiltration,19 and fuel-cell mem* To whom correspondence should be addressed. Address: Department of Chemical and Materials Engineering, University of Kentucky, Lexington, Kentucky 40506-0046. Tel.: 859-257-2794. Fax: 859-3231929. E-mail: [email protected].

brane technology.20 Microporous membranes functionalized with highly reactive metal nanoparticles are quite novel. One application for membrane-supported metal nanoparticles is the catalytic dechlorination of toxic organics.21-27 In a previous study,21,22 we described the reductive dechlorination of trichloroethylene (TCE) and polychlorinated biphenyl (PCB) using Fe/ Pd nanoparticles in poly(acrylic acid) (PAA)/poly(vinylidene fluoride) (PVDF) membranes. PAA (MW 50 000) containing metal ions was added onto the PVDF porous support by a dipcoating method. The metal nanoparticles were then formed by postreduction with sodium borohydride (NaBH4). Because the metal ions were physically mixed with PAA without chelation, this process led to an uncontrollable synthesis of nanoparticles with inhomogeneous dispersion and lower loading. In this study, we took a different approach based on the combination of ion exchange and reduction to prepare metal nanoparticles in membranes. PVDF microfiltration (MF) membranes provide an ideal support for functionalization because of their open structure and large porosity. To fully utilize the pore surface, PAA was synthesized inside the PVDF support membrane by in situ polymerization of acrylic acid. This type of polyelectrolyte pore-filled membrane has been well studied for applications in reverse osmosis (RO), nanofiltration (NF), and pervaporation,17-19 but the in situ synthesis of nanoparticles in the membranes pores has not been reported in the literature. Various metal ions can be introduced into the PAA domain as the nanoparticle precursor by the ion-exchange process. It is also possible to load two or more types of metal ions into the membrane for bimetallic or multimetallic nanoparticle synthesis. In addition to the new synthesis process, the membrane and nanoparticles were fully characterized using various electron microscopy techniques. High-resolution X-ray energy-dispersive spectroscopy (EDS) mapping performed by scanning transmission electron microscopy (STEM) was employed to reveal the elemental distribution at the nanoscale. Our objectives were to synthesize nanostructured Fe/Pd bimetallic particles inside PAA-functionalized PVDF membranes for the reductive dechlorination of toxic organics; study

10.1021/ie0611498 CCC: $37.00 © 2007 American Chemical Society Published on Web 12/13/2006

Ind. Eng. Chem. Res., Vol. 46, No. 8, 2007 2349

Figure 2. Thermally induced free-radical polymerization reaction of acrylic acid using benzoyl peroxide as the initiator. PAA is cross-linked during polymerization by TMPTA.

Figure 1. Schematic diagram of PAA-functionalized PVDF MF membranes prepared using in situ polymerization of acrylic acid via a thermal treatment.

the role of the PAA/PVDF membrane in metal cation binding through the COO- sites, reduction to form nanoparticles, and PCB dechlorination using convective flow under applied pressure; understand and quantify the hydrodechlorination mechanism, the roles of the second dopant metal palladium and of water, and the effects of particle size (nanosize versus bulk) and temperature on the catalytic reactivity. The long-term maintenance of Fe/Pd nanoparticle reactivity was also investigated. Experimental Section Materials. Acrylic acid (AA), benzoyl peroxide, 1,1,1trimethylolpropane triacrylate (TMPTA), toluene anhydrous, ferrous chloride tetrahydrate (FeCl2‚4H2O), potassium tetrachloropalladate (K2PdCl4), sodium borohydride (NaBH4), anhydrous ethanol, and hexane were purchased from Aldrich. Naphthalene-d8 (5000 mg L-1 in methyl chloride), 2,2′dichlorobiphenyl (100 mg L-1 in hexane), 2-chlorobiphenyl (100 mg L-1 in hexane), and biphenyl (100 mg L-1 in hexane) were obtained from Ultra Scientific. Commercial PVDF microfiltration membranes (0.22-µm pore size, 4.7-cm diameter, 125-µm thickness, and 75% porosity) chosen as the substrate for PAA functionalization were obtained from Millipore. Preparation of PAA-Functionalized PVDF Membranes. The PVDF MF membranes functionalized with PAA were prepared by filling the membrane pores with acrylic acid monomer solution and then conducting a literature-reported (for polyethylene MF membranes) in situ free-radical polymerization via a thermal treatment.28 The typical procedure is described in Figure 1. To achieve the ability of wetting out the hydrophobic PVDF membrane, toluene was chosen as the solvent medium. The monomer solution was prepared by mixing acrylic acid (30 wt %), benzoyl peroxide (0.5 wt %, initiator), and TMPTA (1 wt %, cross-linking agent) in toluene. Benzoyl peroxide as the initiator was first dissolved in toluene. The solutions were purged with ultra-high-purity nitrogen for 2 h to remove any dissolved oxygen, which can act as an inhibitor. The PVDF membranes

Figure 3. Schematic diagram of iron nanoparticles synthesized in PAA/ PVDF membranes.

were immersed into the monomer solution for 2 min and quickly placed between two Teflon plates that were subsequently clamped together. The membranes immobilized in the two Teflon plates were then placed into an oven at 90 ( 2 °C with a nitrogen purge. The TMPTA served as the cross-linking monomer because of its trifunctional double bonds (Figure 2). This highly branched structure offers a high cross-linking density during free-radical polymerization.28 After 4 h, the membranes were released from the Teflon plates and washed in 200 mL of ethanol to remove unreacted monomer. Finally, the PAA/PVDF membranes were rinsed with deionized water and kept in deionized water for nanoparticle synthesis. Synthesis of Fe/Pd Nanoparticles in PAA Layers. The flowchart for the membrane-supported nanoparticle synthesis is shown in Figure 3. Briefly, ferrous ions were first loaded into membranes by ion exchange, and Fe nanoparticles were formed followed by reduction of Fe2+ with NaBH4. Prior to the ion exchange, the PAA/PVDF membrane (4.7 cm in diameter) was soaked in 100 mL of sodium hydroxide solution (0.1 M) overnight (12-14 h) to convert PAA from the hydrogen form (-COOH) to the sodium form (-COONa). After the excess sodium hydroxide had been rinsed from the membrane with deionized water, the membrane was shaken in 100 mL of a deoxygenated solution of ferrous chloride (5.5 mM) at pH 4.8-5 with a nitrogen purge for 12 h. During this process, the

2350

Ind. Eng. Chem. Res., Vol. 46, No. 8, 2007

ferrous ions were bound to the PAA in the membrane by ion exchange with Na+. The membrane was then washed with deoxygenated deionized water. Subsequent immersion in 200 mL of sodium borohydride solution (0.07 M) yielded Fe nanoparticles embedded in the PAA/PVDF membrane. It has been reported in the literature that boron is present at ∼4 wt % as FeB in the iron nanoparticles when Fe2+ is reduced by NaBH4.29 The boron and iron contents were determined by digesting the nanoparticles in 25 wt % nitric acid solution and measuring the dissolved Fe and B concentrations through inductively coupled plasma atomic emission spectroscopy (ICPAES). After being rinsed with deoxygenated deionized water and then ethanol, the membrane-supported iron nanoparticles were soaked in 50 mL of a solution (90/10 vol % ethanol/water) of K2PdCl4 (0.12 mM) for 30 min. This resulted in the deposition of Pd on the Fe surface through the following redox reaction26

Pd2+ + Fe0 f Fe2+ + Pd0

(1)

Because of the high reactivity of the Fe nanoparticles, 90 vol % ethanol was used to minimize the Fe corrosion reaction and other side reactions. After being rinsed with ethanol three times, the prepared Fe/Pd nanoparticles in the PAA/PVDF membrane were stored in ethanol solution for further dechlorination studies. The Pd deposition content was also determined by digestion of the nanoparticles in 25 wt % nitric acid, followed by ICP-AES analysis. Fe/Pd nanoparticles with different Pd loadings in the PAA/PVDF membrane were prepared using various K2PdCl4 solutions (0.012-0.2 mM) to investigate the effect of Pd deposition content. Commercial Fe particles were used for Pd coating to compare results with the nanoparticle system. Iron particles (∼120 µm, from Fisher Scientific) were washed with 1 M HCl and then rinsed sequentially with deoxygenated deionized water and ethanol. An aqueous solution of K2PdCl4 was prepared in deoxygenated deionized water (pH 2) and used to deposit Pd on the iron particle surface by the same redox reaction.26 The Fe/Pd particles were then rinsed sequentially with deoxygenated deionized water and ethanol. The rinsed Fe/Pd particles were stored in ethanol solution for the dechlorination study. The Pd coverage was also determined by digestion in 25 wt % nitric acid, followed by ICP-AES analysis. Characterization of Membranes and Fe/Pd Nanoparticles. The surface and cross-sectional morphologies of the unmodified and PAA-functionalized PVDF membranes were examined by scanning electron microscopy (SEM, S-900, Hitachi). Cross sections were prepared by fracturing the membranes in liquid nitrogen. All specimens were coated with a thin layer of gold and palladium. Elemental analysis was performed using a Hitachi S-3200 SEM equipped with an X-ray energy-dispersive spectroscopy (EDS) apparatus. A transmission electron microscope (JEOL JEM-2010F) equipped with an Oxford EDS detector and a scanning (STEM) unit was employed to characterize the membrane nanostructure, i.e., the mean size and size distribution of nanoparticles. The membrane samples for TEM observation were prepared by ultrasectioning the membranes with a diamond knife into slices of about 50-nm thickness using a conventional microtome technique. The cut slices were loaded on a copper TEM grid coated with lacey carbon film. High-resolution EDS mapping analysis was performed in STEM mode to obtain the composition, structure, and distribution of elements at the nanoscale. Details of the STEM-EDS mapping analysis procedure are described in ref 21.

Dechlorination Reactions. Batch experiments for the dechlorination of 2,2′-dichlorobiphenyl (DiCB), a single PCB (polychlorinated biphenyl) congener, were conducted in 24.5-mL serum glass vials. In each batch vial, one piece of PAA/PVDF membrane (47 cm in diameter) embedded with 16 mg of Fe/Pd nanoparticles was loaded into 20 mL of DiCB solution (16 mg L-1 in 50/50 vol % ethanol/water). All of the serum glass vials were sealed with Teflon-lined silicon septa and placed on a wrist-action shaker throughout the duration of the experiment. At predetermined time intervals, a 2-mL aqueous sample was withdrawn from the selected reaction vial and transferred to a 4-mL vial containing 2 mL of hexane for the extraction of PCBs. After removal of all of the residual solution, 10 mL of hexane was added to the reaction vial for membrane phase extraction. The 4-mL vials and the reaction vials were both placed on a wrist-action shaker and mixed for 2 h to achieve extraction equilibrium. Duplicated experiments were conducted at each sampling time. Some experiments were also conducted in convective flow (7-35 bar pressure). Analysis Methods. All PCBs were analyzed by gas chromatograph (Varian-2800) equipped with a mass spectrometer (Varian-6100). From each extraction vial, a 1-mL aliquot of the extraction solvent layer was transferred to a 1-mL GC autosampler vial for analysis by GC/MS. Ten microliters of naphthalene-d8 (5000 mg L-1 in methyl chloride) was added to the GC sample vial as an internal standard (IS). External standards of DiCB, 2-chlorobiphenyl, and biphenyl in hexane were used to prepare calibration curves. The calibration curves for all of the PCBs were linear over the concentration range of 0.5-20 mg L-1 (R2 > 0.999, regressions were based on a sevenpoint calibration). The approximate detection limit for all PCBs was 0.1 mg L-1. According to known sample analysis each time, the maximum error was determined to be less than 10%. Concentrations of all metal ions (iron, sodium, palladium, boron) were quantified by ICP-AES (VISTA-PRO, Varian). In all cases, the instrument calibration was based on commercial standards (Fisher Scientific) containing 1000 mg L-1 of the desired metal serially diluted with 5% nitric acid. Yttrium chloride (1 mg L-1) was used as the internal standard. Judging from the known sample analysis each time, the maximum error for all elements was less than 8%. Results and Discussion Membrane Characterization. Figure 4 compares the SEM surface images of the unmodified PVDF support membrane and the PAA-functionalized PVDF membrane in the dry state. The PVDF MF support membrane (Figure 4a) shows a highly porous microstructure. The pores are mostly circular in shape but highly nonuniform in size (0.2-2 µm). As expected, the modified membrane shows less porosity, with a small number of small pores. This indicates that PAA has been filled into the pores to create smaller pores. As can be seen in Figure 4b, the PVDF support is not completely covered, and the pores are partially blocked after pore modification with 30 wt % acrylic acid in toluene. It has been reported that high pore coverage can be achieved by increasing the concentration of acrylic acid.28 In this study, complete pore filling might not be desirable because of the high diffusion resistance for the hydrophobic chlorinated organic molecules and also for nanoparticle synthesis. The crosssectional images of these membranes are shown in Figure 5. By contrast, different regions of the cross sections clearly show the structural difference between the PAA-modified membrane and the unmodified substrate membrane. As shown in Figure 5c and d, the pores inside the PVDF substrate are filled with


Figure 4. SEM images of membrane surface: (a) unmodified PVDF support membrane, (b) pore-filled PAA/PVDF membrane.

small grains, suggesting that PAA modification has taken place throughout the PVDF substrate. EDS mapping analysis of oxygen atoms was also performed in STEM mode to verify that the observed structural change was due to the PAA modification. The EDS mapping results will be discussed later. Interaction between Ferrous Ions and PAA. By using this method involving a PAA-metal ion binding interaction (Fe2+ and COO-) followed by reduction, we create iron nanoparticles that are well-dispersed inside the membranes because the ferrous ions are bound and distributed along the PAA chains. The surrounding polymer chains also can prevent ion migration and nanoparticle agglomeration, which plays a critical role in stabilizing the nanoparticles and controlling the particle size.15 Therefore, it is necessary to understand the binding interaction between the ferrous ions and PAA. First, the mass balance of sodium and ferrous ions was calculated based on the ICP analysis of the FeCl2 solution before and after the loading of Fe2+ on the membrane. According to the ICP results, the atom ratio of sodium ions released from the PAA/PVDF membrane to the ferrous ions bound to the membrane was 1.9 ( 0.1. In theory, the binding of one ferrous ion from solution results in two sodium ions being released from the membrane because of charge balance. This indicates that the ferrous ions are well chelated with PAA (no physical adsorption). An elemental analysis was performed on the PAA/PVDF membrane loaded with Fe2+, and the results are shown in Figure 6. It has been found that the coordination number for the PAA-divalent metal complex is 2.30 Therefore, one ferrous cation is satisfied with two carboxyl anions containing four oxygen atoms. As shown in Figure 6, the EDS analysis gives the atom ratio of 3.5 (oxygen to iron). This value agrees well with the established PAAmetal binding stability constant. Second, the proximity of the bound ferrous ions in the membrane plays an important role in controlling the size of the nanoparticles.15,22 It has been reported that the average particle size is larger when the metal cation concentrations in the membrane are higher.15,22 This is due to the enhanced aggregation of Fe atoms because of the shorter distance between Fe2+ ions at the higher loading density. To reveal the distributions of Fe and PAA at the nanoscale, EDS elemental maps of the membrane cross sections were acquired by STEM. A region of the membrane sample at lower magnification is shown in an STEM bright-field image in Figure 7a. The EDS mapping was performed at the selected area on the left STEM image (Figure 7a) and reveals the position of atoms of iron, fluorine (F), and

oxygen (O) in Figure 7b-d, respectively. The maps were generated by placing a white dot on the image when an X-ray count of a particular element was received. As shown in Figure 7b and c, the dots for oxygen appear strongly in the map, and oxygen atoms are mainly found in the regions where few fluorine atoms are located. This indicates the presence of PAA inside the membrane because oxygen comes only from carboxylic acid groups. It also confirms the assumption that the small grains observed in the SEM cross sections are PAA. By comparing the iron, fluorine, and oxygen maps, one can see that the iron and oxygen atoms are combined together and located in the same phase separated from the phase where the fluorine atoms are present. This indicates that the PVDF substrate has no affinity for Fe2+ and that the ferrous cations are strongly bound to the PAA carboxylic acid groups. Next, an EDS map was acquired at higher magnification to obtain a better understanding of the interaction between Fe2+ and PAA. As shown in Figure 7e-g, all of the iron atoms are associated with oxygen atoms, and the Fe map matches perfectly with the oxygen map. The black dots appearing in the iron map are believed to be gaps between chelated ferrous ions. It is important to point out that these images and maps were obtained with the membrane in the completely dry state because of the requirements of sample preparation and TEM analysis. This can change the morphology of the membrane because PAA is an extremely swellable polyelectrolyte. Iron Mass Balance in Reduction with NaBH4. According to the ICP analysis results and mass balance, a 95% yield of Fe nanoparticle was achieved after the reduction of bound Fe2+ with NaBH4. This indicates insignificant leaching of the nanoparticles from the membrane phase during the reduction process. According to the ICP analysis, about 4 wt % boron was found in the Fe nanoparticles. Obviously, Fe2+ f Fe0 conversion will lead to Na+ binding with COO- sites. Na analysis by SEM-EDS confirmed this case. It has been reported that the reduction of Fe2+ with borohydride in aqueous solution involves three independent reactions.31 The boron content in the Fe-B nanoparticles is determined by the pH, the addition rate, and the concentration of NaBH4 solution. Fe/Pd Nanoparticle Characterization. TEM analysis at low magnification was performed to verify the formation and distribution of nanoparticles inside the PAA/PVDF membranes. As shown in Figure 8, Fe/Pd nanoparticles of spherical shape were homogeneously dispersed in the PAA phase over the membrane cross section. In contrast, the regions containing no

2352


Figure 5. SEM images of membrane cross sections. Unmodified PVDF support membrane: (a) middle of the cross section, (b) bottom of the cross section. Pore-filled PAA/PVDF membrane: (c) middle of the cross section, (d) bottom of the cross section.

nanoparticles are believed to be the PVDF substrate phase. A statistical analysis of the image yielded an average particle size of 30 nm in diameter, with a standard deviation of 5.7 nm. Using a mean diameter of 30 nm, the external surface area for the nanoparticles was calculated to be ∼25 m2 g-1. EDS analysis was also conducted during the TEM observations using a 2-nm electron beam spot to determine the elements present in the nanoparticles. The composition of the nanoparticles identified in the TEM image was also quantified by EDS (Figure 8c). The Pd content was found to be 1.9 wt %, which is consistent with the previous ICP analysis results. Boron, as a light element at low content, was not detected by EDS because of the low energy sensitivity.32 Next, the nanostructures and elemental distribution of the Fe/ Pd nanoparticles were observed by HRTEM and STEM-EDS mapping. Figure 9 shows an STEM bright-field image and the elemental mapping images of the corresponding areas for Fe and Pd. The probe size used was 1 nm in diameter. The mapping images clearly demonstrate a core/shell structure for the Fe/Pd nanoparticles, with Fe in the core region and Pd in the shell

region. This is as expected because Pd was postreduced by Fe0 and deposited on the iron surface. Catalytic Hydrodechlorination of 2,2′-Dichlorobiphenyl. To investigate the catalytic properties of the Fe/Pd nanoparticles synthesized in PAA/PVDF membranes, we studied the reductive hydrodechlorination of 2,2′-dichlorobiphenyl (DiCB, a single PCB congener) using the bimetallic nanoparticles. PCBs are among the most important chlorinated aromatic compounds that cause severe environmental problems because of their hydrophobic nature and excellent chemical stability. The dechlorination mechanism and kinetic rates were investigated using membrane-supported Fe/Pd nanoparticles. To understand and quantify the role of the second dopant metal, we studied the dechlorination rate as a function of Pd content on Fe, as well as reaction temperature. Kinetic Rates and Mechanism. Figure 10 shows the concentration profiles for the batch reaction of DiCB with Fe/ Pd (Pd ) 2.3 wt %) nanoparticles in PAA/PVDF membranes at 25 °C. The membrane-supported Fe/Pd nanoparticles exhibit an extremely high DiCB degradation rate. More than 90%


Figure 6. SEM-EDS spectra of PAA pore-filled PVDF membranes loaded with ferrous ions.

dechlorination of DiCB with a metal loading of only 0.8 g L-1 was achieved within 2 h. Biphenyl was formed as the dominating product. 2-Chlorobiphenyl (CB) as the chlorinated intermediate was identified only at trace levels. This indicates a direct reductive pathway because of the higher degradation rate for the chlorinated intermediate. The carbon mass balance based on the sum of DiCB, CB, and biphenyl was about 91-95% of the initial amount of DiCB. The 5-9% mass loss is due to the extraction equilibrium. It has been established that PCB dechlorination with Fe/Pd nanoparticles can be described by the following pseudo-first-order reaction kinetics

-

dC ) kobsC ) kSAFmaSC dt

(2)

where C is the DiCB concentration (mg L-1) at time t (h) and aS is the specific surface area of the nanoparticles (m2 g-1). Based on the 30-nm average diameter of the Fe/Pd nanoparticles identified by TEM, aS was calculated to be 25 m2 g-1. Fm is the loading of nanoparticles (g L-1). kSA is the surface-areanormalized rate constant (L h-1 m-2), and kobs is the observed rate constant (h-1). Based on the best linear fit to the experimental data (R2 ) 0.996), kobs and kSA were determined to be 1.36 h-1 and 0.068 L h-1 m-2, respectively. Complete conversion to biphenyl was achieved, and only a trace amount of the intermediate (CB) was detected, which indicates a pathway involving the direct formation of biphenyl for the Fe/ Pd nanoparticle system. To understand the role of particle size in terms of reactivity and reaction pathway shift, we studied the dechlorination of DiCB with bulk Fe particles (∼120 µm) coated with Pd (Pd ) 1.5 wt %). At a very high metal loading of 87.5 g L-1, only about 10% dechlorination of DiCB was achieved, and 2-chlorobiphenyl and biphenyl both appeared at the same concentration level within 8 h (results not shown). The low observed reaction rate was not just due to the lower surface area of the bulk particles (2 m2 g-1 based on BET analysis). However, the surface-area-normalized rate constant kSA calculated for the degradation of DiCB was only 0.00011 L h-1 m-2, which is

over 600 times lower than that obtained for the membranesupported nanosized Fe/Pd particles. The great difference in kSA indicates the higher reactivity of the nanosized Fe/Pd particles. The enhanced reactivity is believed to be due to the various facets, edges, corners, and defects of the nanoparticles, which provide additional sites with high catalytic activities.33 The bulk Fe/Pd system showed a sequential reaction pathway with significant formation of chlorinated intermediates, in contrast to the direct biphenyl formation pathway for the nanosized Fe/ Pd system. Dechlorination of DiCB under Convective Flow. Dechlorination of DiCB by Fe/Pd nanoparticles immobilized on a PAA/ PVDF membrane under convective flow was investigated using a dead-end filtration module supplied by Osmonics. This apparatus has a membrane area of 8 cm2 and contains a stirring device placed in close contact with the feed solution-membrane interface to minimize the effects of concentration polarization. Prior to the dechlorination study, pure water permeation was measured at various pressures to determine the membrane permeability. The unmodified PVDF support membrane had a permeability of 210 × 10-4 cm3 cm-2 bar-1 s-1. After the membrane had been modified with PAA and loaded with Fe/ Pd nanoparticles, the value decreased dramatically to 0.33 × 10-4 cm3 cm-2 bar-1 s-1, and the porosity (inside PAA containing pore), determined by the weight of pure water in membrane pores, decreased from 75% to 58% after PAA and Fe/Pd nanoparticle functionalization. The decreases of the flux and of the porosity are mainly due to the presence of PAA inside the membrane pores, which can be confirmed from the SEM observations. Compared to the 1000-fold reduction in permeability, the membrane porosity reduction is only from 75% to 58%. Because PAA is a strong adsorbent for water, it can enhance the water uptake of the PAA/PVDF membrane despite the decrease of pore volume. In addition to porosity, the water flux also depends on the pore size, geometry, and tortuosity. PAA is synthesized not only inside the membrane pores but also on the surface, which can cause some inside membrane pores to be inaccessible for fluid flow.

2354


Figure 7. STEM-EDS mapping of PAA/PVDF membranes chelated with Fe2+: (a) Low-magnification STEM image of membrane cross section, (b) Fe map from a, (c) F map from a, (d) O map from a, (e) high-magnification STEM image of membrane cross section, (f) Fe map from e, (g) O map from e.

The dechlorination experiments were carried out at four different constant permeation rates of 2.0 × 10-4, 2.6 × 10-4, 5.0 × 10-4, and 11.5 × 10-4 cm3 cm-2 s-1 at steady sate with the initial DiCB concentration of 15.2 mg L-1. The membrane area was 8 cm2, and the membrane contains 7.4 mg of Fe/Pd nanoparticles. Figure 11 shows the PCB concentration profile at different residence times. The residence time was calculated as residence time ) (membrane pore volume)/(permeation rate), and the membrane pore volume was determined from the water uptake. As expected, the conversion of DiCB increased with increasing residence time. The reaction needs a residence time of about 40 s for the complete dechlorination of DiCB. To compare this result with the batch-mode data, kSA under convective flow mode was calculated using the DiCB conversion at a residence time of 6.3 s. By assuming a simple tubular reactor

model, kSA(convective flow) was determined to be 0.064 L h-1 m-2 by the equation

-rA ) kSAFmaSC )

dC dτ

(3)

where C is the DiCB concentration, Fm is the mass concentration of metals in the membrane pore volume (metals/membrane pore volume, g L-1), and τ is the residence time. The kSA value calculated for convective flow mode is close to kSA(batch). This indicates the existence of diffusion control from the nanoparticles located in the pore regions that are inaccessible to fluid flow unless very high pressure is applied. Reaction Mechanism and the Role of Water. To understand the dechlorination mechanism and the role of H2O in DiCB


Figure 8. (a) TEM image of PAA/PVDF membrane cross section containing Fe/Pd nanoparticles, (b) histogram from the left TEM image of 100 Fe/Pd nanoparticles, (c) EDS spectrum acquired from the nanoparticles in the TEM image.

Figure 9. Characterization of Fe/Pd nanoparticles: (a) STEM image of Fe/Pd (Pd ) 2.3 wt %) nanoparticles, (b) EDS mapping image of Fe, (c) EDS mapping image of Pd, (d) high-resolution TEM image of Fe/Pd nanoparticles.

dechlorination with the bimetallic membrane system, DiCB dechlorination in batch mode was performed in the absence of water with membrane-immobilized Fe/Pd nanoparticles. One piece of PAA/PVDF membrane containing 16 mg of Fe/Pd nanoparticles was added into 20 mL of DiCB solution (20 mg L-1 in pure ethanol). No degradation of DiCB or formation of CB and biphenyl was detected after 2 h, indicating that no

dechlorination reaction took place in the absence of water, which is the donor of hydrogen from the Fe corrosion reaction. Subsequently, 10 µL of deionized ultrafiltered (DIUF) water (water concentration ) 27 mM) was deliberately added into the 20-mL DiCB solution. CB and biphenyl were identified after 2 h (Table 1). This indicates the hydrodechlorination reaction requires the presence of some water to facilitate the corrosion

2356


Figure 10. Batch reaction of 2,2′-chlorobiphenyl with Fe/Pd (Pd ) 2.3 wt %) in a PAA/PVDF membrane at room temperature. Metal loading ) 0.8 g L-1. Initial organic concentration ) 16 mg L-1.

Figure 11. Reaction of 2,2′-chlorobiphenyl with Fe/Pd (Pd ) 2.3 wt %) in a PAA/PVDF membrane under convective flow. Initial organic concentration ) 15.2 mg L-1. Membrane area ) 8 cm2 (containing 7.4 mg of Fe/Pd nanoparticles). Membrane thickness ) 125 µm. Operating pressure ) 7-35 bar. Table 1. DiCB Dechlorination after 2 h in Pure Ethanol and 27 mM Watera

pure ethanol 27 mM water in ethanol

DiCB (mM)

CB (mM)

biphenyl (mM)

total (mM)

0.089 0.08

b 0.0015

b 0.0033

0.089 0.085

a Initial DiCB concentration ) 0.089 mM (20 mg L-1). Metal loading ) 16 mg/20 mL, Fe/Pd (Pd ) 2.3 wt %) nanoparticles in PAA/PVDF membrane. b No reaction occurred.

of the primary metal (Fe) to provide hydrogen and electrons that can be utilized to replace chlorine at the Pd (dopant metal) surface. In the bimetallic system, iron serves primarily as an electron donor that provides hydrogen by the reduction of water (corrosion of iron), whereas the second metal (Pd) acts as a catalyst.21,22 The hydrogen gas generated from iron corrosion is adsorbed on the palladium lattice and dissociated with atomic H, which is one of the strongest reductants. It is well-known that chlorinated organics are strongly adsorbed on the palladium surface by forming Pd-Cl bonds.34,35 In the Fe/ Pd bimetallic system, the catalytic hydrodechlorination of dichlorophenol has been reported over the palladium surface by atomic H atoms and electrons provided by Fe-Pd galvanic cells.36

Figure 12. Best linear fit of kSA for the dechlorination of DiCB with various Fe/Pd nanoparticles in PAA/PVDF membranes. Metal loading ) 16 mg/ 20 mL.

Catalytic Activity as a Function of Pd Coating Content. To understand the role of Pd as the second dopant, batch dechlorination rates of DiCB were measured as a function of Pd coating content (Figure 12). kSA values of 0.017, 0.068, and 0.166 L h-1 m-2 were determined for Fe/Pd nanoparticles with 0.6, 2.3, and 5.6 wt % Pd respectively. Once again, in all three of these different Fe/Pd nanoparticle systems, the direct formation of biphenyl was achieved, whereas CB was detected only at trace levels. It should be noted that, at the same batch reaction conditions, Fe0 nanoparticles without Pd showed insignificant dechlorination (less than 0.2 mg L-1 CB was formed after 5 days, and no biphenyl was detected). In the bimetallic system, the role of Fe is to generate hydrogen by the corrosion reaction, whereas Pd serves as the catalyst and the chlorine atom in DiCB is mainly replaced by hydrogen on the Pd surface.22,27 Therefore, the Pd atoms are considered as the surface reactive sites for the dechlorination of DiCB. The variation of kSA as a function of Pd content is due to the difference in reactive sites. By normalizing kSA in terms of Pd content (reactive sites), we found the same reaction rate of Fe/Pd nanoparticles. The following reaction model developed by Johnson provided a better way to understand and quantify the effects of variations in the reactivity of different metal systems21,37

-

dC ) kSAFmaSC ) k2ΓaSFmC dt

(4)

where k2 is the second-order rate constant at a particular type of site (L h-1 mol-1) and Γ is the surface concentration of reactive sites (mol m-2). In this model, kSA is expressed as the product of k2 and Γ, which is more reasonable when the dechlorination reaction preferentially occurs at the reactive catalytic surface sites (bimetallic system). Based on a 30-nm average diameter of nanoparticles, we calculated the Pd coverage and surface Pd atoms for different Fe/Pd nanoparticles using a Pd atom cross-sectional area of 0.0787 nm2.38 Our calculations indicate that Fe/Pd nanoparticles with 0.6, 2.3, and 5.6 wt % Pd have 0.1, 0.4, and 0.97 layers of Pd atoms, respectively. Because the maximum Pd coverage is less than one layer, all of the Pd atoms are considered as surface reactive sites. The Γ values for the three different nanoparticles with 0.6, 2.3, and 5.6 wt % Pd were calculated as 2.20 × 10-6, 8.43 × 10-6, and 2.05 × 10-5 mol m-2, respectively. It should be noted that the total surface area was used in all kSA calculations. By substituting the Γ value into eq 3, k2 was


Figure 14. Best linear fit of Ea for the dechlorination of DiCB with membrane-supported Fe/Pd (Pd ) 2.3 wt %) nanoparticles at various temperatures.

Figure 13. STEM-EDS mapping of Fe/Pd nanoparticles: (a) 0.6 and (b) 5.6 wt % Pd.

determined to be 7727, 8066, and 8098 L h-1 mol-1, respectively. The enhanced reaction rate (kSA) is due only to the increase of the surface Pd atoms. High-resolution STEM-EDS mapping images were also acquired (Figure 13) to compare the Pd atom distribution for different Pd coating nanoparticles. The STEM-EDS mapping technique presents a 2-D image of a 3-D sample in transmission.32 All Fe/Pd nanoparticles showed a core/shell structure with an Fe-rich core region and a Pd-rich shell region. More Pd atoms were deposited on the iron surface and the Pd shell layer became denser with increasing Pd content. In spite of the limited spatial resolution in the EDS mapping, the distribution of Pd atoms is still in qualitative agreement with the results based on the calculations. This result implies that a uniform Pd coating with a controllable thickness can be obtained by postreduction of Pd2+ with Fe nanoparticles immobilized in the membrane phase. In our previous study,22 where commercial PAA was dipcoated on the PVDF surface, we reported a kSA value of 0.10 L h-1 m-2 for Fe/Pd (Pd ) 1 wt %) nanoparticles synthesized on the membrane surface. This is about 9 times higher than the kSA value reported in this article. The higher kSA value is due to the lower mass-transfer resistance because the Fe/Pd nanoparticles were located at the membrane surface. In contrast, in this study, PAA was prepared inside the PVDF membrane pores. Fe2+ was bound to the PAA and reduced to nanoparticles inside the membrane pores. Because the reaction occurs only at the nanoparticle surface, the organics must diffuse through the membrane pores, which can control the reaction rate. However, diffusion control can be avoided by operating at high convective flow. Compared to the previous study (dip-coating), this new synthesis method (pore-filling) provides a more stable membrane

matrix because PAA is cross-linked and immobilized inside the porous network structure instead of physically sitting on top of the membrane surface. Higher PAA and metal loadings can be achieved because the membrane pore surfaces are fully utilized. Effect of Temperature. In an effort to achieve deeper insight into the mechanism of Pd-catalyzed dechlorination, the dependence of the batch reaction rate on temperature was investigated to obtain the activation energy (Ea, kJ mol-1). Ea is the measure of minimum energy required to complete the reaction. In general, the role of the catalyst is to reduce Ea by changing the reaction pathway and thus enhance the reaction rate. It has been found that the complete dechlorination of PCBs in aqueous solution by monometallic bulk Fe0 (no catalyst) requires high temperature (400 °C).39 However, the value of Ea has not been reported in the literature for this type of process. The reduction of PCBs by Ni-Mo catalyst with supplied H2 in a nonaqueous phase was observed at relatively mild conditions (82-91% conversion at 250 °C).40 Earlier research reported the Ea of 124 kJ mol-1 for 2,3-dichlorobiphenyl40 by noncatalytic process and 93 kJ mol-1 for 3-chlorobiphenyl41 by Ni-Mo catalyst with supplied H2 in a nonaqueous phase. The dependence of the rate constant on the temperature is modeled in Figure 14 by the Arrhenius equation. Based on the best linear fitting, the Ea value for DiCB degradation by membrane-supported Fe/Pd nanoparticles was determined to be 24.5 kJ mol-1. This is about 5 times lower than the value for the noncatalytic process40 and nearly 4 times lower than the value for Ni-Mo-catalyzed dechlorination41 in a nonaqueous phase with supplied H2 gas. This indicates the high catalytic activity of Pd, which is effective in the hydrodechlorination of PCBs from water at room temperature. It has been reported in the literature42 that diffusion-controlled reactions in solution have low activation energies (Ea < 21 kJ mol-1), whereas Ea values for reaction on surfaces are usually on the order of 84 kJ mol-1. These Ea values indicate diffusion control for the membrane-supported Fe/Pd system. Stability of Fe/Pd Reactivity. One of the important issues for PCB dechlorination by membrane-supported Fe/Pd nanoparticles is the decrease in the reactivity due to the consumption of iron as a reactant, oxidation of the iron surface, formation of iron(II, III) hydroxide on the particle surface, and possible leaching of Pd or Fe0 from the membrane. An eight-cycle dechlorination experiment was conducted in 20 mL of a batch solution containing 64 mg of membrane-supported Fe/Pd nanoparticles (Pd ) 2.3 wt %) with repeated spiking with a concentrated DiCB solution. After each cycle, the DiCB

2358


concentration was raised to ∼16 mg L-1 by addition of concentrated DiCB solution. More than 90% dechlorination was achieved after the first 4 cycles, whereas only about 50% dechlorination was obtained after 8 cycles, indicating the decline of catalytic activity. It is well-known that Pd and chloride have a strong interaction, which might cause the Pd deactivation. To verify this hypothesis, a dechlorination experiment was conducted in a 40 mg L-1 Cl- solution using fresh membranesupported Fe/Pd nanoparticles. No obvious decline of nanoparticle reactivity was observed. This indicates that Pd is not deactivated by Cl- ions formed from PCB dechlorination. To further investigate the decrease in the catalytic activity, STEMEDS mapping analysis on the nanoparticles (no membrane) was used to examine the distribution of Fe and Pd after 8 cycles. In comparison with fresh Fe/Pd nanoparticles in which Pd was deposited on the Fe surface (Figure 9), it was found that the Pd was completely covered by thick layers of iron and oxygen after the reaction, suggesting the deposition of iron hydroxide, which deactivated the Fe/Pd surface. The final metal ion concentrations in the solution were found to be 0.4 mg L-1 for iron and negligible for Pd (i.e., Pd leaching was not detected), indicating that the Fe/Pd nanoparticles remains stable inside the PAA/ PVDF membrane matrix. Conclusions Fe/Pd nanoparticles were successfully incorporated in PAA/ PVDF membranes by ion exchange and subsequent reduction. The functionalized membranes were characterized using the STEM-EDS mapping technique to distinguish the PAAfunctional domain from the PVDF substrate. High-resolution STEM-EDS mapping analysis was also used to investigate the binding interaction between the metal cations and the carboxylic acid groups, which plays an important role in determining the nanoparticle size and distribution. The Fe/Pd nanoparticle size, structure, and distribution were analyzed by TEM, HRTEM, and STEM-EDS mapping. The membrane-supported nanoparticles exhibit high reactivity in terms of the dechlorination of chlorinated organics. The effect on the catalytic activity of Pd content was studied to understand and quantify the role of Pd in the bimetallic system. The high catalytic activity of Pd was confirmed by the low activation energy compared to those of other catalytic systems. Acknowledgment This study was supported by the NIEHS-SBRP program. We thank Dr. Alan Dozier for assistance with TEM and STEMEDS mapping analysis and John May and Tricia Coakley from the U.K. Environmental Research and Training Laboratory (ERTL) for GC-MS and ICP-AES analytical support. Literature Cited (1) Scott, R. W. J.; Wilson, O. M.; Crooks, R. M. Synthesis, characterization, and applications of dendrimer-encapsulated nanoparticles. J. Phys. Chem. B 2005, 109, 692-704. (2) Jain, P. K.; Lee, K. S.; El-Sayed, I. H.; El-Sayed, M. A. Calculated absorption and scattering properties of gold nanoparticles of different size, shape, and composition: Applications in biological imaging and biomedicine. J. Phys. Chem. B 2006, 110, 7238-7248. (3) Paul, S.; Pearson, C.; Molloy, A.; Cousins, M. A.; Green, M.; Kolliopoulou, S.; Dimitrakis, P.; Normand, P.; Tsoukalas, D.; Petty, M. C. Langmuir-Blodgett film deposition of metallic nanoparticles and their application to electronic memory structures. Nano Lett. 2003, 3, 533-536. (4) Yakushiji, K.; Ernult, F.; Imamura, H.; Yamane, K.; Mitani, S.; Takanashi, K.; Takahashi, S.; Maekawa, S.; Fujimori H. Enhanced spin accumulation and novel magnetotransport in nanoparticles. Nat. Mater. 2005, 4, 57-61.

(5) Patolsky, F.; Weizmann, Y.; Williner I. Actin-based metallic nanowires as bio-nanotransporters. Nat. Mater. 2004, 3, 692-695. (6) Jiang, C.; Markutsya, S.; Pikus, Y.; Tsukruk, V. Freely suspended nanocomposite membranes as highly sensitive sensors. Nat. Mater. 2004, 3, 721-728. (7) Worden, J. G.; Dai, Q.; Shaffer, A. W.; Huo, Q. Monofunctional group-modified gold nanoparticles from solid phase synthesis approach: Solid support and experimental condition effect. Chem. Mater. 2004, 16, 3746-3755. (8) Fukasawa, T.; Suetsuna, T.; Harada, K.; Suenaga, S. Catalytic performance of metal nanoparticles supported by ceramic composite produced by partial reduction of solid solution with dopant. J. Am. Ceram. Soc. 2005, 88, 2938-2941. (9) Liu, Z.; Ling, X. Y.; Su, X.; Lee, J. Y. Carbon-supported Pt and PtRu nanoparticles as catalysts for a direct methanol fuel cell. J. Phys. Chem. B 2004, 108, 8234-8240. (10) Ikeda, S.; Akamatsu, K.; Nawafune, H.; Nishino, T.; Deki, S. Formation and growth of copper nanoparticles from ion-doped precursor polyimide layers. J. Phys. Chem. B 2004, 108, 15599-15607. (11) He, J.; Ichinose, I.; Kunitake, T.; Nakao, A. In situ synthesis of noble metal nanoparticles in ultrathin TiO2-gel films by a combination of ion-exchange and reduction processes. Langmuir 2002, 18, 10005-10010. (12) Pivin, J. C.; Sendova-Vassileva, M.; Lagarde, G.; Singh, F.; Podhorodecki, A. Optical activation of Eu3+ ions by Ag nanoparticles in ion exchanged silica-gel films. J. Phys. D: Appl. Phys. 2006, 39, 29552958. (13) He, J.; Ichinose, I.; Kunitake, T.; Nakao, A.; Shiraishi, Y.; Toshima, N. Facile fabrication of Ag-Pd bimetallic nanoparticles in ultrathin TiO2gel films: Nanoparticle morphology and catalytic activity. J. Am. Chem. Soc. 2003, 125, 11034-11040. (14) Damle, C.; Biswas, K.; Sastry, M. Synthesis of Au-core/Pt-shell nanoparticles within thermally evaporated fatty amine films and their lowtemperature alloying. Langmuir 2001, 17, 7156-7159. (15) Wang, T. C.; Rubner, M. F.; Cohen, R. E. Polyelectrolyte multilayer nanoreactors for preparing silver nanoparticle composites: Controlling metal concentration and nanoparticle size. Langmuir 2002, 18, 3370-3375. (16) Ritchie, S. M. C.; Kissick, K. E.; Bachas, L. G.; Sikdar, S. K.; Parikh, C.; Bhattacharyya, D. Polycysteine and other polyamino acid functionalized microfiltration membranes for heavy metal capture. EnViron. Sci. Technol. 2001, 35, 3252-3258. (17) Yamaguchi, T.; Nakao, S.; Kimura, S. Plasma graft filling polymerization: Preparation of a new type of pervaporation membranes for organic liquid mixtures. Macromolecules 1991, 24, 5522-5527. (18) Mika, A. M.; Childs, R. F.; Dickson, J. M.; McCarry, B. E.; Gagnon, D. R. Porous polyelectrolyte-filled membranes: Effect of cross-linking on flux and separation, J. Membr. Sci. 1997, 135, 81. (19) Zhou, J.; Childs, R. F.; Mika, A. M. Pore-filled nanofiltration membranes based on poly(2-acrylamido-2-methylpropanesulfonic acid) gels. J. Membr. Sci. 2005, 254, 89-99. (20) Nasef, M. M.; Zubir, N. A.; Ismail, A. F.; Khayet, M.; Dahlan, K. Z. M.; Saidi, H.; Rohani, R.; Ngah, T. I. S.; Sulaiman, N. A. PSSA porefilled PVDF membranes by simultaneous electron beam irradiation: Preparation and transport characteristics of protons and methanol. J. Membr. Sci. 2006, 268, 96-108. (21) Xu, J.; Dozier A.; Bhattacharyya, D. Synthesis of nanoscale bimetallic particles in polyelectrolyte membrane matrix for reductive transformation of halogenated organic compounds. J. Nanopart. Res. 2005, 7, 449-467. (22) Xu, J.; Bhattacharyya, D. Membrane-based bimetallic nanoparticles for environmental remediation: Synthesis and reactive properties. EnViron. Prog. 2005, 24, 358-366. (23) Liu, Y.; Choi, H.; Dionysiou, D.; Lowry, G. V. Trichloroethene hydrodechlorination in water by highly disordered monometallic nanoiron. Chem. Mater. 2005, 17, 5315-5322. (24) Jovanovic, G. N.; Znidarsic Plazl, P.; Sakrittichai, P.; Al-Khaldi, K. Dechlorination of p-chlorophenol in a microreactor with bimetallic Pd/ Fe catalyst. Ind. Eng. Chem. Res. 2005, 44, 5099-5106. (25) Tee, Y.-H.; Grulke, E.; Bhattacharyya, D. Role of Ni/Fe nanoparticle composition on the degradation of trichloroethylene from water. Ind. Eng. Chem. Res. 2005, 44, 7062-7070. (26) He, F.; Zhao, D. Preparation and characterization of a new class of starch-stabilized bimetallic nanoparticles for degradation of chlorinated hydrocarbons in water. EnViron. Sci. Technol. 2005, 39, 3314-3320. (27) Lowry, G. V.; Johnson, K. M. Congener-specific dechlorination of dissolved PCBs by microscale and nanoscale zerovalent iron in a water/ methanol solution. EnViron. Sci. Technol. 2004, 38, 5208-5216. (28) Gabriel, E. M.; Gillberg, G. E. In situ modification of microporous membranes. J. Appl. Polym. Sci. 1993, 48, 2081-2090.

Ind. Eng. Chem. Res., Vol. 46, No. 8, 2007 2359 (29) Liu, Y.; Majetich, S. A.; Tilton, R. D.; Sholl, D. S.; Lowry G. V. TCE dechlorination rates, pathways, and efficiency of nanoscale iron particles with different properties. EnViron. Sci. Technol. 2005, 39, 13381345. (30) Tomida, T.; Hamaguchi, K.; Tunashima, S.; Katoh, M.; Masuda, S. Binding properties of a water-soluble chelating polymer with divalent metal ions measured by ultrafiltration. poly(acrylic acid). Ind. Eng. Chem. Res. 2001, 40, 3557-3562. (31) Shen J.; Li, Z.; Yan Q.; Chen, Y. Reactions of bivalent metal ions with borohydride in aqueous soluiton for the preparation of ultrafine amorphous alloy particles. J. Phys. Chem. 1993, 97, 8504-8511. (32) Williams D. B.; Carter, C. B. Transmission Electron Microscopy; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1996. (33) Henry, C.R. Growth, structure and morphology of supported metal clusters studied by surface science techniques. Cryst. Res. Technol. 1998, 33, 1119-1140. (34) Bodnariuk, P.; Coq, B.; Ferrat, G.; Figueras, F. Carbon-chlorine hydrogenolysis over PdRh and PdSn bimetallic catalysts. J. Catal. 1989, 116, 459-466. (35) Park, K. T.; Klier, K.; Wang, C. B.; Zhang, W. X. Interaction of tetrachloroethylene with Pd(100) studied by high-resolution X-ray photoemission spectroscopy. J. Phys. Chem. B 1997, 101, 5420-5428. (36) Wei, J.; Xu, X.; Liu, Y.; Wang, D. Catalytic hydrodechlorination of 2,4-dichlorophenol over nanoscale Pd/Fe: Reaction pathway and some experimental parameters. Water Res. 2006, 40, 348-354.

(37) Johnson, T. L.; Scherer M. M.; Tratnyek, P. G. Kinetics of halogenated organic compound degradation by iron metal. EnViron. Sci. Technol. 1996, 32, 2634-2640. (38) Nutt, M.; Hughes, J. B.; Wong, M. Designing Pd-on-Au bimetallic nanoparticle catalysts for trichloroethene hydrodechlorination. EnViron. Sci. Technol. 2005, 39, 1346-1353. (39) Chuang, F. W.; Larson, R. A.; Wessman, M. S. Zero-valent ironpromoted dechlorination of polychlorinated biphenyls. EnViron. Sci. Technol. 1995, 29, 2460-2463. (40) Gryglewicz, G.; Stolarski, M.; Gryglewicz, S.; Klijanienko, A.; Piechocki, W.; Hoste, S.; Van Driessche, I.; Carleer, R.; Yperman, J. Hydrodechlorination of dichlorobiphenyls over Ni-Mo/Al2O3 catalysts prepared by spray-drying method. Chemosphere 2006, 62, 135-141. (41) Murena, F.; Schioppa, E.; Gioia, F. Catalytic hydrodechlorination of a PCB dielectric oil. EnViron. Sci. Technol. 2000, 34, 4382-4385. (42) Su, C.; Plus, R. W. Kinetics of trichloroethene reduction by zerovalent iron and tin: Pretreatment effect, apparent activation energy, and intermediate products. EnViron. Sci. Technol. 1999, 33, 163-168.

ReceiVed for reView August 31, 2006 ReVised manuscript receiVed October 19, 2006 Accepted October 24, 2006 IE0611498

Pd Nanoparticle Immobilization in Microfiltration Membrane Pores

Recommend Documents