Pu(V)O2+ Adsorption and Reduction by Synthetic Hematite and

Changes in aqueous- and solid-phase plutonium oxidation state were monitored over time in hematite (α-Fe2O3) and goethite (α-FeOOH) suspensions ...
0 downloads 0 Views 384KB Size
Environ. Sci. Technol. 2005, 39, 2107-2114

Pu(V)O2+ Adsorption and Reduction by Synthetic Hematite and Goethite B R I A N A . P O W E L L , * ,† R O B E R T A . F J E L D , † DANIEL I. KAPLAN,‡ JOHN T. COATES,† AND STEVEN M. SERKIZ‡ Department of Environmental Engineering and Science, Clemson University, 342 Computer Court, Anderson, South Carolina 29625, and Westinghouse Savannah River Company, Aiken, South Carolina 29808

Changes in aqueous- and solid-phase plutonium oxidation state were monitored over time in hematite (R-Fe2O3) and goethite (R-FeOOH) suspensions containing 239Pu(V)amended 0.01 M NaCl. Solid-phase oxidation state distribution was quantified by leaching plutonium into the aqueous phase and applying an ultrafiltration/solvent extraction technique. The technique was verified using oxidation state analogues of plutonium and sediment-free controls of known Pu oxidation state. Batch kinetic experiments were conducted at hematite and goethite concentrations between 10 and 500 m2 L-1 in the pH range of 3-8. Surfacemediated reduction of Pu(V) was observed for both minerals at pH values of 4.5 and greater. At pH 3 no adsorption of Pu(V) was observed on either goethite or hematite; consequently, no reduction was observed. For hematite, adsorption of Pu(V) was the rate-limiting step in the adsorption/reduction process. In the pH range of 5-8, the overall removal of Pu(V) from the system (solid and aqueous phases) was found to be approximately second order with respect to hematite concentration and of order -0.39 with respect to the hydrogen ion concentration. The overall reaction rate constant (krxn), including both adsorption and reduction of Pu(V), was 1.75 ( 2.05 × 10-10 (m-2 L)-2.08 (mol-1 L)-0.39 (s-1). In contrast to hematite, Pu(V) adsorption to goethite occurred rapidly relative to reduction. At a given pH, the reduction rate was approximately independent of the goethite concentration, although the hydrogen ion concentration (pH) had only a slight effect on the overall reaction rate. For goethite, the overall reaction rates at pH 5 and pH 8 were 6.0 × 10-5 and 1.5 × 10-4 s-1, respectively. For hematite, the reaction rate increased by 3 orders of magnitude across the same pH range.

(e.g., complexation, precipitation, adsorption, colloid formation, and oxidation/reduction (redox) reactions) (1-4). Pu mobility has been shown to exhibit a very high sensitivity to oxidation state (5-7). Plutonium can exist simultaneously in four oxidation states [Pu(III), Pu(IV), Pu(V), and Pu(VI)] in the natural environment (8). The predominant oxidation states in oxic waters are Pu(IV), Pu(V), and Pu(VI) (9). Pu(IV) generally has a high affinity for the solid phase, yielding sediment distribution coefficients 2 or 3 orders of magnitude greater than Pu(V) (10, 11). Additionally, Pu(V/VI) is appreciably more soluble than Pu(VI), which precipitates as Pu(OH)4(s) or PuO2(s) (12-14). Pu(IV) forms strong hydroxide complexes, even at low pH values, and precipitates as Pu(OH)4(s) in neutral pH solutions (9, 12). The aqueous-phase concentration of Pu(IV) is controlled by formation of plutonium (hydr)oxide solid phases. Typically, solutions with high Eh and pH values tend to favor Pu(V) and Pu(VI) (15). Pu(V) has been shown as the stable aqueous-phase oxidation state in seawater and dilute salt solutions (16, 17). KeeneyKennicutt and Morse (18) speculated that Pu associated with sediments is typically found as Pu(IV) and that soluble Pu is present as Pu(V) or Pu(VI). This is reasonable given the high affinity of Pu(IV) for the solid phase and its low solubility. Previous studies have shown that Pu(IV) and Pu(V) can be sequestered by many environmentally relevant solid phases such as iron oxides (5, 6, 18-21), manganese oxides (19, 21, 23-25), and aluminosilicate minerals (19). Several of these studies have also indicated that Pu undergoes oxidation state transformations in reactions with iron oxide surfaces (5, 6, 18-20). Keeney-Kennicutt and Morse (18) observed simultaneous oxidation and reduction of Pu(V) upon adsorption to goethite followed by the slow reduction of Pu(VI) to Pu(IV). Penrose et al. (6) also observed Pu(V) reduction on goethite and natural sediments. Kersting et al. (19) used X-ray absorption spectroscopy (XAS) to show reduction of Pu(V) by a number of minerals including goethite, birnessite (δ-MnO2), pyrolusite (β-MnO2), and silica. Additionally, reduction of Pu(V) during transport through Idaho National Engineering and Environmental Laboratory interbed sediments and basalt and Savannah River Site (SRS) sediments has been inferred from laboratory column and field lysimeter studies (26-28). This work is a continuation of a study examining the reduction of Pu(V) by a series of iron (oxyhydr)oxide minerals (20). The goal of this work was to develop rate expressions describing the adsorption/ reduction of Pu(V) by synthetic hematite and goethite in the pH range of 3-8. Hematite and goethite were selected for study because they are the most common iron (oxyhydr)oxide minerals found in oxidizing sediments (29), including those of the SRS (30).

Materials and Methods Introduction Many facilities operated by the United States Department of Energy (DOE) require reliable models for predicting the subsurface transport of plutonium and other actinides to evaluate risk posed by subsurface contamination, to design remediation strategies for contaminated sites, and to dispose of Pu-bearing waste safely. These models have the greatest utility if they embody the complex geochemistry of plutonium * Corresponding author present address: Lawrence Berkeley National Laboratory, Chemical Sciences Division, Berkeley, CA; telephone: (510) 486-5377; e-mail: [email protected]. † Clemson University. ‡ Westinghouse Savannah River Company. 10.1021/es0487168 CCC: $30.25 Published on Web 02/15/2005

 2005 American Chemical Society

Hematite and goethite were synthesized using procedures from Schwertmann and Cornell (31). A summary of the physical characteristics of the hematite and goethite used in these studies is listed in the Supporting Information. A description of the techniques used for mineral characterization was previously reported (20). Pu(V) Working Solution Preparation. A fresh Pu(V) working solution was prepared for each experiment by oxidizing an aliquot of a 239Pu(IV)(NO3)4 stock solution as previously described (20). Oxidation state distribution analysis was performed by parallel extraction of the Pu solution into 0.5 M thenoyltrifluoroacetone (TTA; Alfa Asear, Ward Hill, MA) in cyclohexane at pH 0.5 and 0.5 M bis(ethyhexyl)phosphoric acid (HDEHP; Alfa Asear, Ward Hill, MA) in VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2107

heptane at pH 0.5 (32-34). A 0.5 M TTA solution in cyclohexane extracts Pu(IV) from a pH 0 aqueous phase, leaving Pu(V) and Pu(VI) behind. A 0.5 M HDEHP solution in heptane extracts Pu(IV) and Pu(VI) from a pH 0 aqueous phase, leaving Pu(V) behind. The average percent of Pu(V) in the 19 Pu(V) working solutions prepared for these experiments was 96 ( 1.5% with a maximum and minimum of 99% and 93%, respectively. 239Pu concentrations were measured with an R-β discriminating liquid scintillation counter (Wallac Inc., model 1415, Boston, MA). Experimental Conditions and Methods. The experimental design for these adsorption/reduction studies was similar to that previously described for Pu(V) adsorption and reduction by synthetic magnetite (20). A brief description of the kinetic experiments is provided below. Samples were prepared in 15-mL polypropylene centrifuge tubes by adding 10 mL of a 1.5 × 10-8 M Pu(V) working solution at a desired pH to a tube containing a given mass of hematite or goethite to yield the desired surface area concentration. The pH was measured using an Orion Triode pH electrode, calibrated with standard buffers (pH 4-10, Orion) and an Orion 420A meter. Following addition of the Pu(V) working solution, the suspension was mechanically mixed in the dark for various reaction times. The oxidation state distribution of Pu in each sample for each reaction time was measured using a combined ultrafiltration and solvent extraction technique (18, 20, 21). The procedure permits monitoring the oxidation state distribution in the aqueous phase as well as in the total system (solid and aqueous phase) by lowering the pH to leach Pu from the solid phase. It should be noted that this is an indirect measurement of the oxidation state distribution, as Pu must be removed from the solid phase prior to analysis. A detailed description of this oxidation state analysis technique was previously reported (20). A flow diagram of this procedure is listed in the Supporting Information. Briefly, for each reaction time, a 2.5-mL aliquot was removed and filtered at 12 nm using Mircosep 30K MWCO centrifugal filters. An aliquot of the filtrate was removed to determine the aqueous-phase Pu concentration, and oxidation state distribution in the remaining filtrate was measured using the parallel solvent extraction technique discussed above. These data can be used to express the concentration and oxidation state distribution of Pu in the aqueous phase (eq 1):

[Pu]aq ) [Pu(IV)]aq + [Pu(V)]aq + [Pu(VI)]aq

(1)

The pH of the remaining sample was lowered to 1.5 using HClO4 to quantitatively leach Pu(V) and Pu(VI) from the mineral surface. Quantitative leaching of Pu(V) and Pu(VI) was verified using Th(IV), Np(V), and U(VI) as oxidation state analogues for Pu(IV), Pu(V), and Pu(VI), respectively. Incomplete leaching of tetravalent actinides was quantified and accounted for in the determination of the Pu oxidation state distribution by assuming that any Pu remaining on the solid phase was Pu(IV). The Pu mass balance for the total system (solid and aqueous phases) is expressed in eq 2: Pu mTotal ) [Pu(IV)]aqVaq + [Pu(IV)]solidmsolid + [Pu(V)]aqVaq + [Pu(V)]solidmsolid + [Pu(VI)]aqVaq + [Pu(VI)]solidmsolid (2)

where m is the mass of the solid phase (g), V is the volume of the aqueous phase (L), and [Pu(IV)], [Pu(V)], and [Pu(VI)] are the concentrations of Pu in the these oxidation states (mol/L or mol/g). The Pu distribution among the three oxidation states is expressed as a fraction to account for both the solid- and aqueous-phase contributions (eq 3): 2108

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

Pu(Total) Pu(IV) Pu(V) Pu(VI) f aq+solid ) 1.0 ) f aq+solid + f aq+solid + f aq+solid (3)

The stability of Pu(IV), Pu(V), and Pu(VI) throughout the oxidation state analysis procedure was verified by analyzing solid-free, 0.01 M NaCl control samples of Pu(IV), Pu(V), and Pu(VI) (20). Mineral Dissolution and Reduction Potential Measurements. Suspensions of hematite, goethite, and a control containing no solid phase were prepared at pH 3 and pH 8 in 0.01 M NaCl. The hematite and goethite surface area concentrations were 10 m2 L-1. The Eh and iron content of these samples was measured over a 15-day period to monitor for Eh changes or mineral dissolution. An Orion platinum electrode with an Ag/AgCl2 reference (model 96-78) and an Orion 420A meter were used to measure Eh. Aliquots were filtered at 200 nm (Whatman 25 mm nylon syringe filter), and iron concentration in the filtrate was measured using flame atomic absorption (Perkin-Elmer model 5100 spectrophotometer). Validation of Oxidation State Analysis Procedure. The preparation of samples to verify the oxidation state analysis technique using oxidation state analogues was conducted as follows. A small aliquot of 230Th(IV), 237Np(V), or 233U(VI) (from nitric acid stock solutions) was added to 2.0 g L-1 hematite and goethite suspensions in 0.01 M NaCl at pH 8.0 buffered using Na2B4O7‚10H2O. The initial actinide concentrations were [230Th] ) 5.0 × 10-6 M, [237Np] ) 1.4 × 10-7 M, and [233U] ) 1.0 × 10-8 M. The samples were mixed in the dark for a 1-h adsorption step. The samples were then acidified to pH 1.5, leached for various times, and filtered at 12 nm. A 1.0mL aliquot of the filtrate was analyzed by liquid scintillation counting. Light Irradiation Experiments. Previous experiments studying Pu(V) adsorption and reduction on goethite indicated that the reaction could be photochemically catalyzed (5, 18). To study this effect, samples were prepared as described above and mixed on a benchtop shaker table. During mixing the samples were irradiated with a dual-bulb fluorescent lamp (Sylvania Cool White F15T8/CW bulb). The lamp was placed at varying distances from the samples to yield the desired light intensity. Samples were reacted for 30 min each at pH 6.5. The light intensity was measured using a Traceable Dual-Display Light meter (Control Company, Friendswood, TX). Following the reaction period, the samples were analyzed using the combined ultrafiltration and solvent extraction technique discussed above. Rate Constant and Reaction Order Determination. A detailed description of the methods by which adsorption and reaction rate constants were determined was previously described (20). Pseduo-first-order adsorption rate constants (k′ads) were used to describe the adsorption of Pu to the solid phase and were based upon measurement of the aqueous Pu concentration. The rate constants were obtained from

d[Pu]aq ) -k′ads[Pu]aq dt

(4)

where [Pu]aq is the concentration of Pu in the aqueous phase (eq 1). Since the solution of eq 4 is exponential with time, the value of the slope of ln[Pu]aq/[Pu]aq,initial versus time is the pseudo-first-order adsorption rate constant. During these studies, Pu(V) was the initial oxidation state, and no other oxidation state was measured in the aqueous phase during the course of the experiments. Therefore, the rate constants obtained from eq 4 are Pu(V) adsorption rate constants. In a few cases, adsorption occurred too rapidly to obtain a reliable rate constant. In these cases, the rate constants are reported as greater than the value of the slope between zero and the first sampling interval. These rate constants were not used in any further calculation or comparisons.

Pseudo-first-order overall reaction rate constants (k′rxn) were also determined. They include both the adsorption and the reduction of Pu(V) in the system based upon the results from the oxidation state analysis procedure described above for the total system (solid and aqueous phases). Equation 5 describes the adsorption and reduction of Pu in the total system as a function of mineral surface area concentration [MSA], hydrogen ion concentration [H+], fraction of Pu(V) in Pu(V) the system f aq+solid , and the overall reaction rate constant krxn: Pu(V) df aq+solid Pu(V) ) -krxnf aq+solid [MSA]m[H+]n dt

(5)

The coefficients m and n are the reaction orders for mineral and hydrogen ion concentrations, respectively. The mineral surface area concentration of hematite and goethite are noted as [R-Fe2O3] and [R-FeOOH], respectively. Errors reported for oxidation state distributions were propagated from Pu counting statistics. All errors reported for rate constants and reaction order terms represent the standard deviation obtained from linear regression reported at the 95% confidence level. The pH, [R-FeOOH], and [R-Fe2O3] reported for each set of experiments are the average from all samples for a given experiment. Errors reported for pH, [R-FeOOH], and [R-Fe2O3] of each experiment are twice the standard deviation of the average.

Results and Discussion Experimental and Pu Oxidation State Analysis Controls. The oxidation state of Pu(V) working solutions (with no solid phase present) at pH 3, 5, and 8 were monitored over 30 days. These solutions showed no change in Pu(V) concentration. Additionally, oxidation state pure Pu(IV), Pu(V), and Pu(VI) solutions remained as 98 ( 4% Pu(IV), 97 ( 4% Pu(V), and 98 ( 2% Pu(VI) after being carried through the ultrafiltration/solvent extraction oxidation state analysis technique described above. The integrity of the initial oxidation state was maintained throughout the procedure, validating the stability of each oxidation state as well as the separation method in 0.01 M NaCl. Therefore, no changes in the aqueous Pu oxidation states occurred as experimental artifacts. Several solutions were prepared to observe the possibility of mineral dissolution and Eh changes occurring during experiments. The average Eh of pH 3 solutions of hematite, goethite, and 0.01 M NaCl were 864 ( 70, 862 ( 79, and 845 ( 82 mV, respectively. The average Eh of pH 8 solutions of hematite, goethite, and 0.01 M NaCl were 521 ( 25, 524 ( 29, and 535 ( 37 mV, respectively. Thus, there was little change in Eh due to the addition of hematite or goethite. Additionally, iron was not measured above the detection limit of 0.1 mg L-1 in the aqueous phase at any pH. Results from adsorption and leaching experiments using Th(IV), Np(V), and U(VI) as oxidation state analogues for Pu(IV), Pu(V), and Pu(VI), respectively, are shown in Figure 1 for hematite and goethite. After 1-h contact time at pH 8, there was considerable adsorption of each actinide, shown as the adsorption step. Actinide concentrations in control samples did not change, indicating negligible loss to container walls. After lowering the pH to 1.5, the actinides were leached for 15 min (indicated as the leaching step in Figure 1). Data for additional leaching times are presented in the Supporting Information. These data indicate that, for both hematite and goethite, a 15-min leaching time is adequate to quantitatively leach Np(V) and U(VI) into the aqueous phase. Additionally, a significant fraction of Th(IV) was leached from the goethite within 15 min. However, Lu et al. (22) and Powell et al. (20) showed that the fraction of leachable Th(IV)/Pu(IV) from iron oxides decreased over time.

FIGURE 1. Adsorption and leaching of (a) 230Th, (b) 237Np, and (c) 233U from hematite and goethite. Control does not contain hematite or goethite. Adsorption step 1 h at pH 8; leaching step 15 min at pH 1.5. [230Th] ) 5.0 × 10-6 M, [237Np] ) 1.4 × 10-7 M, and [233U] ) 1.0 × 10-8 M.

FIGURE 2. Adsorption of Pu(V) to hematite vs time. It is desirable to keep the leaching time as short as possible to minimize unintentionally oxidizing Pu during sample processing. These experiments indicate that Np(V) and U(VI) can be quantitatively leached from hematite and goethite after mixing the suspension for 15 min at pH 1.5. Quantitative desorption of Pu(IV)/Th(IV) is not necessary for this analysis based upon the mass balance assumption in eq 2. Pu(V) Adsorption to Hematite and Goethite. Adsorption of Pu(V) to hematite and goethite over time is shown in Figures 2 and 3, respectively. The rate of adsorption of Pu(V) to hematite increased with increasing pH and with increasing hematite concentration. In solutions with low carbonate concentrations, Pu(V) remains as the free dioxycation and VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2109

FIGURE 3. Adsorption of Pu(V) to goethite vs time. does not begin hydrolyzing until a pH between 9 and 10 is reached. Therefore within the pH range studied here, PuO2+ is the predominant aqueous-phase species. As the pH increases, the hematite surface becomes more negatively charged, causing a greater attraction of the cationic PuO2+ species and yielding faster adsorption. The increasing adsorption rate with increasing hematite concentration is attributed to a greater probability of interaction of PuO2+ with a hematite adsorption site. These systems are not expected to exhibit site-limited kinetics. The available mineral surface area in the 10 m2 L-1 solutions was estimated to be 3 orders of magnitude greater than the area that could be occupied by the available Pu (assuming that an adsorbed Pu atom/complex would take up 1 nm2 surface area). No adsorption of Pu(V) onto hematite or goethite occurred in pH 3 solutions. Effects of pH and mineral concentration were far more subtle for goethite than for hematite. Increasing pH and increasing goethite concentration increased the rate of Pu(V) adsorption, as observed for hematite. However, Pu(V) adsorption to goethite was very rapid for all pH values and mineral concentrations when compared to hematite systems of the same pH and mineral concentration. This effect was more pronounced for pH 5 and pH 6.5 systems. Around pH 8, the adsorption rates were comparable, as >90% of the Pu(V) was adsorbed within 30 min for both minerals. However, around pH 6, >90% Pu(V) adsorption occurred after 360 min in goethite solutions and ∼4300 min in hematite solutions of comparable mineral concentrations. Pu(V) Reduction by Hematite. The kinetics of Pu(V) reduction by hematite were studied in 10-100 m2 L-1 hematite solutions from pH 3 to pH 8. At pH 3, no adsorption or reduction of Pu(V) was observed in samples monitored over 30 days. In contrast, reduction was observed in all hematite systems at pH g 4.5. The total system oxidation state distribution of Pu in 10 m2 L-1 solutions at pH 6.09 ( 0.02 and pH 7.24 ( 0.03 are shown in Figure 4. The steady decline in Pu(V) and the corresponding increase in Pu(IV) imply reduction of Pu(V) to Pu(IV). Importantly, measurement of Pu in the aqueous phase showed only Pu(V), verifying that the observed reduction was not taking place in the aqueous phase (data not shown). Because aqueous-phase reduction was not observed and no reduction was observed in pH 3 systems, where no adsorption occurred, it is inferred that reduction of Pu(V) is mediated by the hematite surface. Pseudo-first-order overall reaction rates, k′rxn, of 7.50 ( 0.68 × 10-6 and 2.00 ( 0.16 × 10-5 s-1 were calculated for the pH 6.09 and pH 7.24 systems, respectively (Table 1). The overall reaction rate increased with increasing pH, presumably the result of an increase in the adsorption rate as discussed above. The fraction of each Pu oxidation state on the solid phase was inferred by subtracting the aqueous-phase Pu oxidation state distribution from the total distribution. This was done by rearranging eq 2 for [Pu(IV)]solidmsolid, [Pu(V)]solidmsolid, or [Pu(VI)]solidmsolid and normalizing the data by dividing each term by the total mass of Pu added. Solid-phase oxidation 2110

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

FIGURE 4. Total system Pu oxidation state distribution at (a) pH 6.09 ( 0.02 and (b) pH 7.24 ( 0.03. [r-Fe2O3] ) 10 m2 L-1; 0.01 M NaCl; initially added [Pu(V)aq] ) 1.5 × 10-8 M. Error bars represent a 95% confidence level based upon error propagation of counting statistics.

TABLE 1. Adsorption and Reduction Rate Constants for Pu(V) Interaction with Hematite

pH

[r-Fe2O3] m2 L-1

apparent adsorption rate constant kads ′ (s-1)a

3.03 ( 0.02 9.8 ( 0.3 no adsorption 3.00 ( 0.08 100.8 ( 1.5 no adsorption 4.60 ( 0.05 10.0 ( 0.3 8.70 ( 0.63 × 10-7 4.90 ( 0.07 9.9 ( 0.3 2.30 ( 0.18 × 10-6 6.09 ( 0.04 10.2 ( 0.3 8.67 ( 1.20 × 10-6 7.24 ( 0.03 9.9 ( 0.3 2.20 ( 0.31 × 10-5 7.76 ( 0.02 49.5 ( 1.6 1.08 ( 0.11 × 10-3 7.90 ( 0.02 99.3 ( 3.0 4.30 ( 0.55 × 10-3

apparent reaction rate constant krxn ′ (s-1)b no reduction no reduction 8.30 ( 0.38 × 10-7 2.40 ( 0.38 × 10-6 7.50 ( 0.68 × 10-6 2.00 ( 0.16 × 10-5 7.60 ( 0.16 × 10-4 2.22 ( 0.29 × 10-3

a From measured aqueous Pu concentration data (eq 4). b From pseudo-first-order approximation of measured aqueous and solid-phase Pu(V) concentration data (eq 5).

state distribution is shown in Figure 5 for the same hematite systems discussed above. From these data, it is clear that Pu(IV) was the dominant solid-phase Pu species. These data indicate that reduction was very rapid following adsorption, thus making adsorption a rate-limiting step. Similar behavior was observed for all hematite concentrations in the pH range of 5-8. The same rate limitation was observed for the reduction of Pu(V) by synthetic magnetite (17). The pseudofirst-order adsorption and overall reaction constants for all pH values and hematite concentrations are presented in Table 1. It is seen that k′rxn and k′ads are of the same order of magnitude in all systems with low pH and hematite concentrations. The similarity between k′ads and k′rxn in all cases further demonstrates that adsorption is the rate-limiting step. The rate-limiting adsorption step hinders calculation of an independent reduction rate constant. However, the overall reaction rate expressions in Table 1 are useful because they describe the combined adsorption and reduction of Pu(V) by hematite. There is difference between k′ads and k′rxn in two systems with high pH values and the highest hematite concentrations. Because these systems are at high pH and hematite concentrations, they are expected to exhibit the

FIGURE 7. Dependence of log k′rxn on log[H+]. [H+] reaction order term, n, calculated via eq 5; n ) -0.39 ( 0.05; krxn ) 1.75 ( 2.05 × 10-10 (m-2 L)-2.08 (mol-1 L)-0.39 (s-1); R2 ) 0.967. Each data point is generated from approximately 10 observations. increases with increasing pH. Since adsorption is the ratelimiting step in these systems, this increasing reaction rate represents the increased attraction of PuO2+ as the hematite surface becomes more negative with increasing pH. The intercept of the line shown in Figure 7 yielded an overall reaction rate constant krxn. Combining the data discussed above, the reduction of Pu(V) by synthetic hematite in the pH range of 4.6-7.9 can be described by FIGURE 5. Solid-phase Pu oxidation state distribution on hematite at (a) pH 6.09 ( 0.02 and (b) pH 7.74 ( 0.05. [r-Fe2O3] ) 10 m2 L-1; 0.01 M NaCl; added [Pu(V)aq] ) 1.5 × 10-8 M.

FIGURE 6. Dependence of log k′rxn on log[r-Fe2O3]. [r-Fe2O3] reaction order term, m, based on eq 5; m ) 2.08 ( 0.32, R2 ) 0.994. Each data point is generated from approximately 10 observations. fastest adsorption kinetics. The difference between k′ads and k′rxn suggests that the rate-limiting effects of Pu adsorption are less subtle in these systems. The adsorption rate constants were calculated based upon the loss of Pu from the aqueous phase using eq 4, and the overall reaction rate constants were based upon the loss of Pu(V) from the system using eq 5. Thus, these two rate constants were calculated from different data sets and can be compared without bias. A plot of k′rxn versus hematite concentration for the three data sets near pH 8 is shown in Figure 6. The slope of the line in Figure 6 represents the hematite reaction order m (eq 5). The slope was found to be 2.08 ( 0.32, indicating approximately second-order kinetics with respect to the hematite concentration. The calculated hematite reaction order term was used to normalize all of the pseudo-firstorder overall reaction rates to the respective hematite concentration. These normalized rate constants are plotted against log [H+] in Figure 7. The hydrogen ion reaction order term n was -0.39 ( 0.05 (eq 5), indicating a relatively weak influence of pH on the reaction rate. The hematite reaction order, in terms of [H+], measured here is similar to the value of -0.34 ( 0.02 measured for Pu reduction by magnetite in previous studies (20). The negative value of the hydrogen ion reaction order term indicates that the reaction rate

Pu(V) df aq+solid Pu(V) ) -krxnf aq+solid [R-Fe2O3]2.08(0.32[H+]-0.39(0.05 dt (6)

with krxn ) 1.75 ( 2.05 × 10-10 (m-2 L)-2.08 (mol-1 L)-0.39 (s-1). This rate is 2 orders of magnitude lower than that measured for Pu(V) adsorption and reduction by synthetic magnetite (20). Reduction of Pu(V) by Goethite. The kinetics of Pu(V) reduction by goethite were studied in 20-500 m2 L-1 goethite solutions in the pH range of 3-8. The total system oxidation state distribution of Pu in goethite solutions at pH 6.47 ( 0.02 and pH 7.96 ( 0.05 is shown in Figure 8. As with hematite, there was a decrease in the Pu(V) fraction in the goethite system with a corresponding increase in the Pu(IV) fraction, and no adsorption or reduction of Pu(V) was observed in systems at pH 3 (Figure 3). Additionally, no Pu(IV) was observed in the aqueous phase, regardless of pH (data not shown). Thus, it can be concluded that reduction is surface mediated by both goethite and hematite. A significant fraction of the Pu adsorbed to goethite was Pu(V), indicating that adsorption may not be a rate-limiting step for Pu(V) reduction by goethite (Figure 9). This is likely due to the relatively fast adsorption kinetics for Pu(V) on goethite as compared to hematite. The solid-phase oxidation state distribution was found by subtracting the aqueousphase Pu distribution from the total system data shown in Figure 8. At pH 7.95, most of the Pu(V) was rapidly adsorbed to goethite and then slowly reduced to Pu(IV) (Figure 9b). Similar results were obtained at pH 6.47, although adsorption of Pu(V) to goethite was slower. The influence of pH and mineral concentration on Pu(V) reduction by goethite are far more subtle than for hematite. Table 2 lists the pseudo-first-order adsorption rate constant, k′ads, and pseudo-first-order overall reaction rate constant, k′rxn, obtained for each system. Goethite concentrations of 20.9, 51.0, and 100.0 m2 L-1 near pH 8 yielded similar overall reaction rates. The same effect was observed for systems around pH 6.5 for goethite concentrations ranging from 20 to 500 m2 L-1. Experiments were run at goethite concentrations of 300 and 500 m2 L-1 in an effort to favor more rapid adsorption. In both cases >90% adsorption occurred within the first 15 min and was too rapid to calculate an adsorption rate. However, there was no significant difference in the VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2111

TABLE 2. Adsorption and Reduction Rate Constants for Pu(V) Interaction with Goethite

pH

[r-FeOOH] m2L-1

apparent adsorption rate constant k′ads (s-1)a

2.94 ( 0.06 5.69 ( 0.29 6.47 ( 0.04 6.47 ( 0.02 6.66 ( 0.16 6.54 ( 0.03 6.56 ( 0.02 8.09 ( 0.12 7.96 ( 0.05 7.75 ( 0.13

50.0 ( 1.2 101.5 ( 1.5 20.5 ( 2.2 48.3 ( 4.1 101.1 ( 1.6 300.5 ( 7.6 502.2 ( 7.1 20.9 ( 0.9 51.0 ( 2.2 100.0 ( 1.6

no adsorption 8.5 ( 0.37 × 10-5 9.1 ( 0.9 × 10-5 9.0 ( 1.60 × 10-5 8.1 ( 0.65 × 10-5 >2.0 × 10-3 c >3.0 × 10-3 c 2.5 ( 0.37 × 10-4 >1.0 × 10-3 c >5.0 × 10-2 c

apparent reaction rate constant k′rxn (s-1)b no reduction 4.7 ( 0.5 × 10-5 6.2 ( 1.1 × 10-5 7.2 ( 1.6 × 10-5 5.8 ( 1.7 × 10-5 6.3 ( 1.1 × 10-5 7.7 ( 1.4 × 10-5 1.3 ( 0.4 × 10-4 1.3 ( 0.3 × 10-4 1.4 ( 0.4 × 10-4

a See Table 1. b See Table 1. c >90% adsorption observed in