Lead Sorption on Ruthenium Oxide: A Macroscopic and Spectroscopic

Desorption studies as a function of aging time (1 h to 1 year) using a continuous stirred-flow reactor with a background electrolyte (0.01 M NaNO3, pH...
2 downloads 0 Views 146KB Size
Environ. Sci. Technol. 2004, 38, 2836-2842

Lead Sorption on Ruthenium Oxide: A Macroscopic and Spectroscopic Study KIRK G. SCHECKEL,* CHRISTOPHER A. IMPELLITTERI, AND JAMES A. RYAN United States Environmental Protection Agency, ORD, NRMRL, LRPCD, WMB, 5995 Center Hill Avenue, Cincinnati, Ohio 45224

The sorption and desorption of Pb on RuO2‚xH2O were examined kinetically and thermodynamically via spectroscopic and macroscopic investigations. X-ray absorption spectroscopy (XAS) was employed to determine the sorption mechanism with regard to identity and interaction of nearest atomic neighbors, bond distances (R), and coordination numbers (N). The kinetics of the Pb-Ruoxide sorption reaction are rapid with the equilibrium loading of Pb on the surface achieving approximately 1:1 wt/wt (129 µmol m-2). XAS data indicate that Pb adsorbed as bidentate innersphere complexes with first shell Pb-O parameters of RPb-O ) 2.27 Å and NPb-O ) 2.1-2.5. PbRu interatomic associations suggest two distinct bidentate surface coordinations of Pb to edges (RPb-RuI ∼3.38 Å, NPb-RuI ∼1.0) and shared corners (RPb-RuII ∼4.19 Å, NPb-RuII ∼0.8) on RuO2 octahedra (cassiterite-like structure), and an additional second neighbor backscattering of Pb indicates the formation of Pb-Pb dimers (RPb-Pb ∼3.89 Å, NPb-Pb ∼0.9). Desorption studies as a function of aging time (1 h to 1 year) using a continuous stirred-flow reactor with a background electrolyte (0.01 M NaNO3, pH 6) demonstrated that Pb was tightly bound (99.7-99.9% retained). The Pb sorption capacity and retention on RuO2‚xH2O is greater than that of other metal oxides examined in the literature. The results of this study imply that RuO2‚xH2O may serve as a high capacity remediation treatment media.

Introduction Metal oxide phases play an important role in governing the sorption and desorption mechanisms of metals in water, soils, and sediments. Many researchers have examined the efficiency of Pb sorption on Mn (1-7), Fe (2, 3, 8-20), Al (7, 21-28), Ti (29, 30), and Si (31-37) oxide surfaces. Most of these studies concluded that the primary Pb sorption mechanism was adsorption; however, some studies observed induced coprecipitation of Pb with the oxide phase (10, 15, 38). Further, Pb sorption capacity has been extensively studied on a variety of sorbents such as activated carbon (39-43), agricultural byproducts (44-46), cation-exchange resins (4749), and aquatic exoskeletons (50, 51). To be an effective remediation sorbent for Pb removal from solution, a cost analysis for the amount of Pb sorbed per unit cost of sorbent must be determined. Commonly, most of the examples listed previously rarely sorb more than a few weight percent Pb * Corresponding author [email protected].

phone:

(513)487-2865;

e-mail:

2836 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

per unit weight of sorbent; however, they are less expensive to manufacture than RuO2‚xH2O. Beyond the ability to sorb large amounts of Pb from solution, understanding the sorption mechanism and potential stability of the Pb complex yields data applicable to designing higher capacity and more efficient sorbents. Speciation of the adsorbed Pb complex can be best accomplished by X-ray absorption spectroscopy studies. Bargar et al. (22) employed X-ray absorption fine structure (XAFS) spectroscopy to investigate the speciation of Pb complexes sorbed to hematite and goethite. They determined that Pb was adsorbed via a mononuclear bidentate mechanism to the Fe octahedra of the hematite and goethite surfaces regardless of pH, sorption density, and initial Pb concentration. The highest surface loading achieved by Bargar et al. (22) was for Pb sorption ([Pb]o ) 9.69 mM, 9.9 µmol m-2) on hematite at pH 7 resulting in a Pb/hematite ratio of 10% (wt. basis) (percent weight values presented within this current study were calculated from experimental parameters and data presented in the cited literature). Al oxides have also been examined in similar fashion (21, 25). Strawn et al. (25) observed the formation of innersphere bidentate bonding of Pb to γ-Al2O3 at pH 6.5 with a resulting sorption capacity that was approximately 2.6% (1.27 µmol m-2). Bargar et al. (21) studied Pb reactions with R-Al2O3 (with some γ-Al2O3 impurity) using XAFS to determine Pb binding as mono- and polynuclear bidentate complexes on the edges of the Al octahedra. At a solution pH of about 7 yielding the highest surface coverage, Bargar et al. (21) found a Pb sorption maximum of 1.6% (5.2 µmol m-2) on R-Al2O3 resulting in the formation of Pb-Pb dimers. Results from an XAFS investigation of Pb sorption onto amorphous SiO2 by Elzinga and Sparks (52) showed mononuclear innersphere Pb sorption complexes were most prevalent when the pH was less than 4.5. Between 4.5 < pH < 5.6, they observed the formation of surface-attached mononuclear and covalent polynuclear Pb species, possibly Pb-Pb dimers, and these dimers were the dominate sorption product above pH 6. At the higher pH values where Pb-Pb dimers were the predicted sorption complex, the Pb capacity on the amorphous SiO2 surface reached 6% (0.1 µmol m-2). A detailed equilibrium and kinetic study coupled with XAFS analysis for Pb sorption shows again the formation of mononuclear bidentate sorption complexes for Pb on birnessite (pH 3.7) and manganite (pH 6.7) (6). Matocha et al. (6) determined surface loadings of Pb at roughly 28% (3.3 µmol m-2) and 1.3% (0.8 µmol m-2) for birnessite and manganite, respectively. While most spectroscopic studies of Pb sorption to metal oxide surfaces resulted in mononuclear or polynuclear bidentate sorption mechanisms, the type of sorbent significantly influenced the retention of Pb during subsequent desorption analyses. For Pb sorbed to γ-Al2O3 at pH 6.5 (25), desorption employing a replenishment method coupled with a cation-exchange resin with the sorption background electrolyte at pH 6.5 found that the sorbed Pb was 98% reversible. However, Pb removal via the background electrolyte from birnessite (6) resulted in only a fraction of a percent desorbed with respect to the total Pb sorbed. Numerous studies have shown the effect of sorbent type on the retention of Pb on natural materials (9, 15, 16, 53-58) with results varying from 0 to 100% Pb retention. RuO2 has been utilized in catalysis for the production of chlorine (59) and in the electrooxidation of methanol in fuel cells (60). The presence of water disrupts the threedimensional (cassiterite-like) structure of RuO2‚xH2O causing 10.1021/es035212l Not subject to U.S. copyright. Publ. 2004 Am. Chem.Soc. Published on Web 04/06/2004

the structure to become more chainlike resulting in a net increase of reactive sites (60). Impellitteri et al. (61) recently examined the effectiveness of RuO2‚xH2O for the removal of arsenate and arsenite from solution (pH 7) with equilibrium surface loadings approaching 7 and 18 wt %/wt, respectively. Further, Impellitteri et al. (61) demonstrated that a 250 mg/L As(III) solution was completely oxidized within 5 s upon reaction with the RuO2‚xH2O surface followed by rapid adsorption of the As(V) oxidation product. For an anion reacting with a metal oxide surface, these results are quite amazing. Drawing upon the work of Impellitteri et al. (61), we designed experiments to investigate the sorption of Pb on RuO2 via spectroscopic and macroscopic methods to elucidate the sorption mechanism and sorption capacity as well as Pb retention via stirred-flow dissolution.

Experimental Procedures Materials. All chemicals used were ACS reagent grade (Fisher Scientific, Fair Lawn, NJ) unless otherwise noted. Stock solutions of Pb(NO3)2 (1000 and 10 000 mg/L) were made in a background electrolyte of 0.01 M NaNO3. Pb stock solutions were prepared minutes prior to addition to the RuO2‚xH2O suspension and the concentration verified during metal analysis by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) (IRIS Intrepid, ThermoElemental, Franklin, MA). The RuO2‚xH2O material (Alfa Aesar, Ward Hill, MA) was characterized with a surface area of 38 ( 0.7 m2 g-1, PZC ≈ 4.0, and mean particle size of 6.15 µm. X-ray diffraction analysis showed an amorphous material (max. count intensity 150 counts s-1) with two broad peaks at approximately 30-45 2°Θ and 45 to 70 2°Θ. Kinetic Experiments. To investigate the influence of time on Pb sorption, pH-stat (DL77, Mettler-Toledo, Columbus, OH) kinetic experiments were conducted for RuO2‚xH2O reacted with Pb at pH 6. In a 1 L Teflon reaction vessel, 5 g of RuO2‚xH2O was reacted with 500 mL of combined background electrolyte and appropriate stock solution to obtain initial Pb concentrations of 207.2, 2072, and 10 360 mg L-1. These initial concentrations were employed to examine the maximum sorption capacity of Pb on RuO2‚ xH2O. The mixtures were vigorously stirred and purged with N2 to inhibit Pb carbonate formation. Kinetic samples were collected at time 0 through approximately 40 h at which time the reaction appeared to be complete. Aliquots (10 mL) from the reaction vessel were filtered through 0.2 µm nylon filters, acidified, and refrigerated until analysis by ICP-AES. For the long-term desorption studies, similar sorption experiments were designed on a slightly larger level (total volume of 900 mL) at an initial Pb concentration of 10 360 mg L-1 with the first week of reaction under pH-stat followed by periodic pH adjustments coupled with constant agitation by tumbling up to 1 year. The desorption studies (aging times of 1 h, 1 day, 1 week, 1 month, 3 months, 6 months, and 1 year) employed a continuous-stirred flow reactor system (reactor volume ) 60 mL, Soil Measurement Systems, Tucson, AZ) by injecting the background electrolyte (0.01 M NaNO3, pH 6) at a rate of 12 mL/h for 72 h. The stirred flow vessel was equipped with a membrane filter to contain the Pb reacted RuO2‚xH2O but would allow passage of aqueous elements. To the stirred-flow reactor, 60 mL from the longterm kinetic reaction vessel were added and sealed. Calculations were noted to take into account dilution effects of base addition to maintain a solution pH of 6. Flow of the background electrolyte began immediately in which 12 mL of solution was collected each hour in 15 mL test tubes placed within a programmed fraction collector (Spectrum Chromatography, Houston, TX). The collected samples were acidified and refrigerated until analysis by ICP-AES. Batch Sorption Experiments. In triplicate, a 0.02 g sample of RuO2‚xH2O was placed in a 50 mL round-bottom Teflon

tube with the appropriate volume of background electrolyte (pH 6) and stock solution for a total volume of 20 mL (1 g RuO2‚xH2O/1 L of solution). For the equilibrium studies at 22 °C, initial concentrations of Pb increased at 100 mg L-1 increments from 0 to 1000 mg L-1 and 500 mg L-1 increments from 1000 to 2000 mg L-1. The suspensions were placed on an end-over-end shaker at 30 rpm, and the pH (target pH ) 6) of the samples was adjusted 3-5 times daily under N2 purge. The pH was adjusted to pH 6 via inputs of either NaOH or HNO3 until the pH stabilized (no change over a 24 h period) at which point the suspension was tumbled for an additional 24 h for a total reaction time of 5 days. The samples were then filtered through 0.2 µm nylon filters, acidified, and refrigerated until analysis by ICP-AES. The data, after correction for acid/base additions, were analyzed using Freundlich and Langmuir models. X-ray Absorption Spectroscopy. Pb-reacted RuO2‚xH2O kinetic samples were analyzed by X-ray absorption near edge (XANES) and X-ray absorption fine structure (XAFS) spectroscopies to determine Pb speciation and related local atomic parameters for Pb on the surface of RuO2‚xH2O. For the data presented in this paper, we show the results of the 10 360 mg L-1 ([Pb]o) with the assumption that the results are identical to the lower initial Pb concentrations. For XANES and XAFS studies, a thin layer of a Pb-reacted RuO2‚xH2O sample was smeared onto Kapton tape and folded back on itself. Pb (13055 eV) LIII-XANES and XAFS data were collected at beamline 20-BM (Pacific Northwest Consortium-Collaborative Access Team (PNC-CAT)) at the Advanced Photon Source at Argonne National Laboratory, Argonne, IL. The electron storage ring operated at 7 GeV. Three to five scans were collected at ambient temperature in transmission mode. A 0.5 mm premonochromator slit width and a Si(III) double crystal monochromator detuned by 20% to reject higherorder harmonics was employed. The beam energy was calibrated by assigning the first inflection of the absorption edge of lead metal foil to 13 055 eV. XANES and XAFS spectra were collected in both transmission and fluorescence modes with a solid-state 13-element detector. Transmission mode was chosen over fluorescence to avoid conflict of selfabsorption due to the high sorption of Pb on RuO2‚xH2O. Reference samples of Pb4OH44+ and Pb(NO3)2 were collected for comparison with the XANES spectra. The collected scans for a particular sample were averaged, the data were then normalized, and the background was removed by spline fitting using WinXAS 2.0 (62). The XAFS data were then converted to k-space, windowed, and Fourier transformed to convert to R-space. Conventional shell-by-shell fitting of the radial structure functions (RSFs) was attempted by theoretical paths for Pb and O backscatter atoms generated from crystallographic data of model compounds using FEFF 7.0. Variables obtained from the fits include the energy phase shift (∆Eo), coordination numbers (N), bond distances (R), and Debye-Waller factors (σ2) that were derived from nonlinear least-squares fitting. Values for R are accurate from approximately (0.02 Å (first shell) to (0.04 Å (second shell), and N values are generally accurate to (1 (63, 64).

Results and Discussion Kinetic Experiments. Figure 1 shows the kinetic uptake of Pb from solution for initial Pb concentrations of 207.2, 2072, and 10 360 mg L-1 in a 10 g RuO2‚xH2O/1 L suspension. Lead in the lower initial Pb concentrations of 207.2 and 2072 mg L-1 is almost completely sorbed within 5 min of reaction time for sorption weight percents of about 2 and 20%, respectively. It was not until we attempted an extremely high initial Pb concentration of 10 360 mg L-1 under controlled conditions that a measured maximum sorption capacity was obtained of approximately 95% (by weight) of Pb on the RuO2‚ xH2O surface through the 42 h reaction time resulting in a VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2837

FIGURE 1. Kinetics of Pb sorption on RuO2‚xH2O as a function of Pb concentration at pH 6 (I ) 0.01 M NaNO3).

FIGURE 3. Thermodynamic modeling and parameters of Pb sorption on RuO2‚xH2O derived from data in Figure 2.

FIGURE 2. Thermodynamic equilibrium study of Pb sorption on RuO2‚xH2O as a function of Pb concentration at pH 6 (I ) 0.01 M NaNO3, T ) 22 °C). surface coverage of 128.8 µmol m-2. These results demonstrate that Pb sorption on RuO2‚xH2O is very rapid with an enormously high capacity. Analysis of the kinetic data to ordered rate models was not attempted due to the fact that obtaining true data on initial rates by conventional sampling methods was not feasible. Preliminary research studies using pressure jump relaxation kinetics measurement techniques show great promise in ascertaining the kinetic rates of cation and anion sorption on RuO2‚xH2O. Batch Sorption Experiments. The equilibrium sorption results at pH 6 for Pb-RuO2‚xH2O are shown in Figure 2. The initial Pb concentration for the samples ranged from 0 to 2000 mg L-1 for the 1 g RuO2‚xH2O/1 L solution. For nearly all samples, except the three highest Pb concentrations of 1000, 1500, and 2000 mg L-1, no Pb (below the detection limit of 0.014 mg L-1) remained in solution at equilibrium after 5 days of reaction. Despite the tremendous surface coverage of Pb, there is no sharp upward tailing at the higher concentrations that could be indicative of surface precipitation; however, assigning a sorption mechanism based solely on the shape of an adsorption curve should be avoided (65). These data were plotted against the Langmuir and Freundlich adsorption models (Figure 3). The data fit both models well. The Langmuir parameters suggest a maximum monolayer 2838

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

sorption level, b, to be 9.92 × 105 mg kg-1, which is very close to our measured value of approximately 1.00 × 106 mg kg-1. The binding strength constant, k, is 4.65 × 1011 suggesting a strong binding potential. The Freundlich model predicts a distribution coefficient, Kd, of 3.45 × 1013 and a correction factor, n, of 25.9 (1/n ) 0.0386) indicates the nonlinearity of the data. X-ray Absorption Spectroscopy. X-ray absorption near edge and X-ray absorption fine structure spectroscopic spectra from Pb(II) sorbed RuO2‚xH2O samples and reference compounds are shown in Figures 4-7. The PbIII-XANES spectra presented in Figures 4 and 5 are characteristic of the multiple scattering of electrons among neighboring atoms and are sensitive to the first shell coordination milieu that provides information on the molecular structure of the sorbed Pb(II) species on the surface of RuO2‚xH2O. Evaluation of the Pb/RuO2‚xH2O sorption spectra relative to the reference compounds show that small levels of distortion and quantity of primary nearest neighbors (in this case oxygen) in the local environment of Pb(II) results in significant differences in the XANES spectra. The chemical structure of Pb(II) in Pb4(OH)4(aq)4+ possesses a distorted trigonal pyramidal coordination environment with surface oxygen atoms (when sorbed) and available hydroxide ions as coordinating ligands rather than water (21, 66). As such, this Pb(II) chemical environment results in covalent bonding of Pb(II) sorption to surfaces that have been attributed to Pb innersphere adsorption complexes on Fe, Al, Mn, and Si oxides (6, 22, 25, 52) for environmentally relevant reaction conditions. The coordination environment of aqueous Pb(II) as a Pb(NO3)2 basic solution is more indicative of low pH sorption systems

FIGURE 4. Normalized XANES spectra from Pb-reacted RuO2‚xH2O aged samples and reference samples.

FIGURE 6. The k3-weighted χ functions of Pb-reacted RuO2‚xH2O samples aged from 5 min to 1 year.

FIGURE 5. Normalized first derivatives of XANES spectra from Pbreacted RuO2‚xH2O aged samples and reference samples. that promotes a highly ionic Pb(II) species with first shell O characteristics of H2O ligands (52, 67, 68). Relative to the featureless Pb4(OH)4(aq)4+ reference spectra, the Pb(NO3)2 XANES curve has more distinct features as noted by the narrower white-line edge and amplitudes beyond the edge. The XANES spectra for the Pb(II) sorbed RuO2‚xH2O samples are essentially identical to the XANES spectrum of the Pb4(OH)4(aq)4+ reference compound with distorted trigonal pyramidal coordination (Figures 4 and 5). There are no evident similarities between the Pb(II) sorbed RuO2‚xH2O samples and the Pb(NO3)2 based on XANES analyses aside from confirmation of the Pb(II) oxidation state. The differences in the Pb4(OH)4(aq)4+ and Pb(NO3)2 XANES and first derivative of the XANES spectra in comparison to the Pb/ RuO2‚xH2O sorption spectra indicates that Pb(II) sorption complexes on the RuO2‚xH2O surface are unlike Pb(II) atoms with water ligands. Since the Pb/RuO2‚xH2O spectra are similar to the distorted trigonal pyramidal coordination of Pb4(OH)4(aq)4+, one can conclude that covalent bonding of Pb(II) sorption complexes to RuO2‚xH2O surfaces is the principal coordination environment (21). The k3-weighted χ functions collected from XAFS analyses of the Pb(II) sorbed RuO2‚xH2O samples and reference compounds are shown in Figure 6. The χ structures of the Pb(II) sorbed RuO2‚xH2O samples resemble the Pb4(OH)4(aq)4+ reference sample as expected regarding the previous discussion of the XANES data for these samples. As aging time of the kinetic samples increases, there is no apparent change

FIGURE 7. Fourier transformed radial structure functions (RSFs) of PbLIII XAFS spectra for Pb-reacted RuO2‚xH2O aged samples as a function of time. in the k3-weighted χ functions suggesting that the speciation of Pb(II) does not change significantly over time. The k3-weighted χ functions were Fourier transformed (uncorrected for phase shift) to isolate the spectral frequencies within the χ spectra for Pb(II) sorbed RuO2‚xH2O samples and reference compounds (Figure 7). The XAFS fitting results are presented in Table 1 for the Pb(II) sorbed RuO2‚xH2O samples and reference compounds. Reasonable fits were achieved by fitting four shells: Pb-O, two Pb-Ru, and PbPb shells. The radial structure functions (RSFs) in Figure 7 are dominated by one distinct peak as a result of first shell oxygen backscattering between 1.2 and 2.1 Å indicating backscattering from 2.1 to 2.5 oxygen atoms at approximately 2.27 Å for the Pb(II) sorbed RuO2‚xH2O kinetic samples. The Pb-O bond distance (R) and coordination number (N) values are similar to values determined for both reference compounds (R ) 2.25-2.32 Å, N ) 2.0-2.3), although the XANES and k3-weighted χ functions indicate that the Pb/RuO2‚xH2O sorption complexes are covalently bonded as in the Pb4(OH)4(aq)4+ coordination environment. Literature studies have reported that RPb-O values are distinctly different for innerand outersphere sorption mechanisms. Outersphere sorption complexes have been characterized with an approximately 0.2 Å greater RPb-O (2.47-2.53 Å) (52, 67, 68) in comparison VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2839

TABLE 1. Structural Parameters Derived from XAFS Analysis Pb-O

Pb-RuI

Pb-RuII

Pb-Pb

aging times

∆Eo (eV)

R (Å)a

Nb

σ2 (Å2)c

R (Å)a

Nb

σ2 (Å2)c

R (Å)a

Nb

σ2 (Å2)c

R (Å)a

Nb

σ2 (Å2)c

1 year 6 month s 3 month s 1 month 1 week 1 day 1 hour 30 min 5 min Pb4(OH)44+ Pb(NO3)2

1.93 2.01 1.89 1.82 1.75 1.95 1.85 1.78 1.85 3.20 2.59

2.27 2.27 2.27 2.27 2.27 2.27 2.27 2.27 2.27 2.32 2.25

2.5 2.3 2.3 2.2 2.3 2.2 2.1 2.2 2.1 2.0 2.3

0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1

3.38 3.38 3.38 3.37 3.38 3.37 3.37 3.36 3.38 3.12

1.0 0.8 0.7 0.9 0.7 0.7 0.8 0.7 0.6 0.7d

0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

4.19 4.19 4.20 4.19 4.19 4.18 4.19 4.20 4.18

0.9 0.9 0.8 0.8 0.7 0.8 0.8 0.8 0.7

0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1 0.0 1

3.89 3.89 3.90 3.89 3.89 3.88 3.89 3.90 3.88 3.77 3.77

0.5 0.5 0.4 0.4 0.4 0.4 0.4 0.5 0.2 2.2 2.6

0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

a

Interatomic distance.

b

Coordination number. c Debye-Waller Factor (fixed to 0.01).

to RPb-O for innersphere complexes (2.18-2.30 Å) (6, 21, 22, 25, 52) possessing distorted trigonal pyramidal O ligand coordination. The Pb-O bond distance (2.27 Å) for the Pb(II) sorbed RuO2‚xH2O samples suggests the formation of innersphere sorption complexes. Data for Pb-Ru interatomic backscattering derived from the RSFs are presented in Table 1, which shows RPb-Ru at two distinct distances of approximately 3.38 (RPb-RuI) and 4.19 Å (RPb-RuII) with NPb-Ru about 1 in both cases. The Pb-O (from above) and Pb-Ru values along with literature knowledge of innersphere Pb-O bonds and the structure of RuO2 octahedra can be employed to determine more definitively the sorption structure of Pb on RuO2‚xH2O. The RuO2 octahedra are characterized by the ditetragonal dipyramidal space group P42/mnm (69). As such, within a RuO2 octahedron, the Ru atom is bonded to four oxygens in a rectangular radial plane at 2.01 Å with 79 or 101° separations between oxygen atoms in the plane and two 90° axial oxygens are found at 1.92 Å. Distance separations between nearest oxygen atoms in a RuO2 octahedron range from 2.55 to 3.10 Å. Applying this information, one can determine possible sorption geometries of Pb on RuO2 octahedra, Pb-Ru interatomic bond distances, and the number of ways (n) these structures can exist. Four geometric configurations are theoretically possible for Pb(II) bonded to RuO2‚xH2O that include corner-sharing monodentate (RPb-Ru ) 4.24-4.44 Å, n ) 2), edge sharing bidentate (RPb-Ru ) 2.93-3.43 Å, n ) 3), corner-sharing bridging bidentate (RPb-Ru ) 3.93-4.25 Å, n ) 3), and face-sharing tridentate (RPb-Ru ) 2.51-3.25 Å, n ) 2). Of these four possible geometries and taking into account the experimentally determined Pb-O distance (2.27 Å) and coordination number (∼2.2), Pb(II) sorption on RuO2‚xH2O may observe edge-sharing bidentate or cornersharing bridging bidentate complexes. The corner-sharing monodentate and face-sharing tridentate geometries do not fit into the experimental results due to NPb-O values (∼2.2) for both and likely cationic repulsion between Pb(II) and Ru(IV) based on ionic radii values for face-sharing tridentate complexes (21). Experimentally derived RPb-RuI (∼3.38 Å) distances are in line with calculated values associated with edge-sharing bidentate bonding of Pb(II) to RuO2 octahedra (2.93-3.43 Å). Additionally, Pb-Ru interatomic distances for cornersharing bridging bidentate complexes (3.93-4.25 Å) encompass the XAFS determined RPb-RuII (∼4.19 Å) for a second Pb-Ru sorption geometry. These results indicate Pb(II) sorbs as an innersphere species (RPb-O ) 2.27 Å) with at least two surface oxy-ligands (O or OH) either on a shared edge of one octahedron crystal or shared corners of neighboring RuO2 octahedra. There is no evidence of corner-sharing monodentate and face-sharing tridentate geometries as dismissed earlier based on experimental and theoretical constraints. 2840

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

d

Oxygen shell.

FIGURE 8. Rates of Pb desorption from Pb-reacted RuO2‚xH2O samples via continuous stirred-flow desorption as a function of aging time. Evidence for polynuclear Pb(II) adsorption complexes was observed as a third second-neighbor frequency within the XAFS data with good reasonable fits. This backscattering effect corresponds to interactions from neighboring Pb atoms with RPb-Pb ranging from 3.88 to 3.90 Å and is indicative of polynuclear Pb(II) surface complexes, possibly Pb-Pb dimers (21, 52). Calculated RPb-Pb values based on potential polynuclear Pb(II) complex formation of edge-sharing bidentate sorbed Pb on RuO2 octahedra ranges from 3.83 to 4.13 Å. Polynuclear Pb(II) surface complexes are not geometrically or chemically possible between neighboring Pb(II) innersphere corner-sharing bridging bidentate complexes on RuO2 octahedra. Combining the short first neighbor Pb-O and the presence of Pb-Pb bonds indicates probable innersphere, covalent polynuclear Pb-Pb complexes in line with similar literature studies of metal sorption to oxide minerals (13, 70-77). The weak frequency amplitude corresponding to Pb-Pb backscattering suggests that the Pb-Pb complexes are small and do not involve extensive three-dimensional precipitation or formation of neo-silicate-like sheets. With the significant uptake of Pb onto the RuO2‚xH2O surface, the Pb-Pb scattering indicating Pb-Pb dimers was likely enhanced by initial Pb(II) innersphere edge-sharing bidentate complexes acting as apparent nucleation sites for continued Pb removal from solution (52). Pb Retention on RuO2‚xH2O. Pb desorption from RuO2‚ xH2O was characterized by an initial rapid (although minute) release followed by an equilibrium plateau resulting in limited Pb release (Figure 8) during the stirred-flow desorption studies at pH 6. The curves of Pb release in Figure 8 show a maximum desorption of approximately 0.3% for the 1 h aged sample. The 1 year sample exhibited only 0.1% Pb release

over the 72 h reaction period upon reaction with the background electrolyte solution. These data clearly show strong retention of Pb by the RuO2‚xH2O surface at pH 6. However, further studies examining the influence of pH, particularly acidic pH values, need to be assessed to evaluate the feasibility of regenerating the RuO2‚xH2O material for reuse as a sorbent media.

Acknowledgments The U.S. EPA has not subjected this manuscript to internal policy review. Therefore, the research results presented herein do not, necessarily, reflect Agency policy. Mention of trade names of commercial products and companies does not constitute endorsement or recommendation for use. PNCCAT facilities and research at these facilities is supported by the U.S. DOE Office of Science Grant DE-FG03-97ER45628. Use of the Advanced Photon Source was supported by the U.S. DOE, under Contract W-31-109-ENG-38. The authors extend their deepest appreciation to the staff of the Pacific Northwest Consortium-Collaborative Access Team, especially beamline scientist Robert Gordon for his assistance and expertise during data collection. We also thank M. Morrison and W. Shuster for sorbent characterization. The authors appreciate the time and helpful comments of anonymous reviewers.

Literature Cited (1) Hem, J. D. Chem. Geol. 1978, 21, 199-218. (2) McKenzie, R. M. Aust. J. Soil Res. 1980, 18, 61-73. (3) Manceau, A.; Charlet, L.; Boisset, M. C.; Didier, B.; Spadini, L. Appl. Clay. Sci. 1992, 7, 201-223. (4) Dong, D. M.; Nelson, Y. M.; Lion, L. W.; Shuler, M. L.; Ghiorse, W. C. Water Res. 2000, 34, 427-436. (5) Hettiarachchi, G. M.; Pierzynski, G. M.; Ransom, M. D. Environ. Sci. Technol. 2000, 34, 4614-4619. (6) Matocha, C. J.; Elzinga, E. J.; Sparks, D. L. Environ. Sci. Technol. 2001, 35, 2967-2972. (7) Trivedi, P.; Axe, L. Environ. Sci. Technol. 2001, 35, 1779-1784. (8) Benjamin, M. M.; Leckie, J. O. J. Colloid Interface Sci. 1981, 79, 209-221. (9) Eick, M. J.; Peak, J. D.; Brady, P. V.; Pesek, J. D. Soil Sci. 1999, 164, 28-39. (10) Ford, R. G.; Kemner, K. M.; Bertsch, P. M. Geochim. Cosmochim. Acta 1999, 63, 39-48. (11) Martinez, C. E.; McBride, M. B. Environ. Sci. Technol. 1999, 33, 745-750. (12) Christophi, C. A.; Axe, L. J. Environ. Eng. 2000, 126, 66-74. (13) Ostergren, J. D.; Brown, G. E., Jr.; Parks, G. A.; Persson, P. J. Colloid Interface Sci. 2000, 225, 483-493. (14) Elzinga, E. J.; Peak, D.; Sparks, D. L. Geochim. Cosmochim. Acta 2001, 65, 2219-2230. (15) Martinez, C. E.; McBride, M. B. Environ. Toxicol. Chem. 2001, 20, 122-126. (16) Scheinost, A. C.; Abend, S.; Pandya, K. I.; Sparks, D. L. Environ. Sci. Technol. 2001, 35, 1090-1096. (17) Trivedi, P.; Axe, L.; Dyer, J. Colloid Surf. A 2001, 191, 107-121. (18) Manju, G. N.; Anoop Krishnan, K.; Vinod, V. P.; Anirudhan, T. S. J. Hazard. Mater. 2002, 91, 221-238. (19) Dyer, J. A.; Trivedi, P.; Scrivner, N. C.; Sparks, D. L. Environ. Sci. Technol. 2003, 37, 915-922. (20) Grossl, P. R.; Sparks, D. L.; Ainsworth, C. C. Environ. Sci. Technol. 1994, 28, 1422-1429. (21) Bargar, J. R.; Brown, G. E., Jr.; Parks, G. A. Geochim. Cosmochim. Acta 1997, 61, 2617-2637. (22) Bargar, J. R.; Brown, G. E., Jr.; Parks, G. A. Geochim. Cosmochim. Acta 1997, 61, 2639-2652. (23) Bargar, J. R.; Towle, S. N.; Brown, G. E., Jr.; Parks, G. A. J. Colloid Interface Sci. 1997, 185, 473-492. (24) Bargar, J. R.; Brown, G. E., Jr.; Parks, G. A. Geochim. Cosmochim. Acta 1998, 62, 193-207. (25) Strawn, D. G.; Scheidegger, A. M.; Sparks, D. L. Environ. Sci. Technol. 1998, 32, 2596-2601. (26) Shen, S. Y.; Taylor, R. W.; Bart, H.; Tu, S. I. Commun. Soil Sci. Plant Anal. 1999, 30, 2711-2730. (27) Boily, J. F.; Fein, J. B. Chem. Geol. 2000, 168, 239-253.

(28) Lee, G.; Bigham, J. M.; Faure, G. Appl. Geochem. 2002, 17, 569581. (29) Shubha, K. P.; Raji, C.; Anirudhan, T. S. Indian J. Eng. Mater. Sci. 1998, 5, 65-71. (30) Shubha, K. P.; Raji, C.; Anirudhan, T. S. Water Res. 2001, 35, 300-310. (31) Chlopecka, A.; Adriano, D. C. Sci. Total Environ. 1997, 207, 195206. (32) Xu, C. Y.; Schwartz, F. W.; Traina, S. J. Environ. Eng. Sci. 1997, 14, 141-152. (33) Nagata, N.; Kubota, L. T.; Bueno, M.; Peralta-Zamora, P. G. J. Colloid Interface Sci. 1998, 200, 121-125. (34) O’Day, P. A.; Carroll, S. A.; Waychunas, G. A. Environ. Sci. Technol. 1998, 32, 943-955. (35) Moulin, I.; Stone, W. E. E.; Sanz, J.; Bottero, J. Y.; Mosnier, F.; Haehnel, C. Langmuir 1999, 15, 2829-2835. (36) Xu, Y. M.; Wang, R. S.; Wu, F. J. Colloid Interface Sci. 1999, 209, 380-385. (37) Rose, J.; Moulin, I.; Hazemann, J. L.; Masion, A.; Bertsch, P. M.; Bottero, J. Y.; Mosnier, F.; Haehnel, C. Langmuir 2000, 16, 99009906. (38) Ainsworth, C. C.; Pilou, J. L.; Gassman, P. L.; Sluys, W. G. V. D. Soil Sci. Soc. Am. J. 1994, 58, 1615-1623. (39) Krishnan, K. A.; Sheela, A.; Anirudhan, T. S. J. Chem. Technol. Biotechnol. 2003, 78, 642-653. (40) Pesavento, M.; Profumo, A.; Alberti, G.; Conti, F. Anal. Chem. Acta 2003, 480, 171-180. (41) Kannan, N.; Rajakumar, A. Fresenius Environ. Bull. 2002, 11, 160-164. (42) Reed, B. E.; Vaughan, R.; Jiang, L. Q. J. Environ. Eng. 2000, 126, 869-873. (43) Vaughan, R. L.; Reed, B. E.; Viadero, R. C.; Jamil, M.; Berg, M. Adv. Environ. Res. 1999, 3, 229-242. (44) Rao, M.; Parwate, A. V.; Bhole, A. G.; Kadu, P. A. J. Water Supply Res. Technol. 2003, 52, 49-58. (45) Daifullah, A. A. M.; Girgis, B. S.; Gad, H. M. H. Materials Lett. 2003, 57, 1723-1731. (46) Kadirvelu, K.; Namasivayam, C. Environ. Technol. 2000, 21, 1091-1097. (47) Kumar, M.; Rathore, D. P. S.; Singh, A. K. Analyst 2000, 125, 1221-1226. (48) Juang, R. S.; Wang, S. W.; Lin, L. C. J. Membr. Sci. 1999, 160, 225-233. (49) Walcarius, A.; Lamdaouar, A. M.; El Kacemi, K.; Marouf, B.; Bessiere, J. Langmuir 2001, 17, 2258-2264. (50) An, H. K.; Park, B. Y.; Kim, D. S. Water Res. 2001, 35, 3551-3556. (51) Lee, M. Y.; Lee, S. H.; Shin, H. J.; Kajiuchi, T.; Yang, J. W. Process Biochem. 1994, 33, 749-753. (52) Elzinga, E. J.; Sparks, D. L. Envion. Sci. Technol. 2002, 36, 43524357. (53) Bunzl, K.; Schmidt, W.; Sansoni, B. J. Soil Sci. 1976, 27, 32-41. (54) Puls, R. W.; Powell, R. M.; Clark, D.; Eldred, C. J. Water Air Soil Poll. 1991, 57-58, 423-430. (55) Coughlin, B. R.; Stone, A. T. Environ. Sci. Technol. 1995, 29, 2445-2455. (56) Sparks, D. L. J. Plant Nutr. Soil Sci. 2000, 163, 563-570. (57) Strawn, D. G.; Sparks, D. L. Soil Sci. Soc. Am. J. 2000, 64, 144156. (58) Glover, L. J.; Eick, M. J.; Brady, P. V. Soil Sci. Soc. Am. J. 2002, 66, 797-804. (59) Lo, I. M. C.; Yang, X. Y. Water Air Soil Poll. 1999, 109, 219-236. (60) McKeown, D. A.; Hagans, P. L.; Carette, L. P. L.; Russell, A. E.; Swider, K. E.; Rolison, D. R. J. Phys. Chem. B 1999, 103, 48254832. (61) Impellitteri, C. A.; Scheckel, K. G.; Ryan, J. A. Envion. Sci. Technol. 2003, 37, 2936-2940. (62) Ressler, T. J. Synchrotron Rad. 1998, 5, 118-122. (63) O’Day, P. A.; Rehr, J. J.; Zabinsky, S. I.; Brown, G. E., Jr. J. Am. Chem. Soc. 1994, 116, 2938-2949. (64) Morin, G.; Ostergren, J. D.; Juillot, F.; Ildefonse, P.; Calas, G.; Brown, G. E., Jr. Am. Miner. 1999, 84, 420-434. (65) Sparks, D. L. Environmental Soil Chemistry; Academic Press: San Diego, CA, 1995. (66) Hong, S.; Olin, A° . Acta. Chem. Scand. 1974, A28, 233-238. (67) Bargar, J. R.; Towle, S. N.; Brown, G. E., Jr.; Parks, G. A. Geochim. Cosmochim. Acta 1996, 60, 3541-3547. (68) Strawn, D. G.; Sparks, D. L. J. Colloid Interface Sci. 1999, 216, 257-269. (69) Wyckoff, R. W. G. Crystal structures; John Wiley & Sons: New York, 1963; Vol. 1. (70) Charlet, L.; Manceau, A. Geochim. Cosmoshim. Acta 1994, 58, 2577-2582. VOL. 38, NO. 10, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2841

(71) Cheah, S.-F.; Brown, G. E., Jr.; Parks, G. A. Geochim. Cosmochim. Acta 1999, 63, 3229-3246. (72) Cheah, S. F.; Brown, G. E., Jr.; Parks, G. A. J. Colloid Interface Sci. 1998, 208, 110-198. (73) Fendorf, S. E.; Lamble, G. M.; Stapleton, M. G.; Kelley, M. J.; Sparks, D. L. Environ. Sci. Technol. 1994, 28, 284-289. (74) Ostergren, J. D.; Trainor, T. P.; Bargar, J. R.; Brown, G. E., Jr.; Parks, G. A. J. Colloid Interface Sci. 2000, 225, 466-482. (75) Scheinost, A. C.; Ford, R. G.; Sparks, D. L. Geochim. Cosmochim. Acta 1999, 63, 3193-3203.

2842

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 10, 2004

(76) Xia, K.; Mehadi, A.; Taylor, R. W.; Bleam, W. F. J. Colloid Interface Sci. 1997, 185, 252-257. (77) Xia, K.; Taylor, R. W.; Bleam, W. F.; Helmke, P. A. J. Colloid Interface Sci. 1998, 199, 77-82.

Received for review October 30, 2003. Revised manuscript received February 12, 2004. Accepted February 25, 2004. ES035212L