U(VI)-Kaolinite Surface Complexation in Absence and Presence of

HA has no influence on positions of the fluorescence emission bands. ... kinetics, thermodynamics, redox behavior, and speciation of actinides in this...
0 downloads 0 Views 221KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 6142-6147

U(VI)-Kaolinite Surface Complexation in Absence and Presence of Humic Acid Studied by TRLFS ADE ´ LA KR ˇ EPELOVA ´ ,* VINZENZ BRENDLER, SUSANNE SACHS, NILS BAUMANN, AND GERT BERNHARD Forschungszentrum Dresden-Rossendorf e.V., Institute of Radiochemistry, P.O. Box 510119, D-01314 Dresden, Germany

Time-resolved laser-induced fluorescence spectroscopy (TRLFS) was applied to study the U(VI) surface complexes on kaolinite in the presence and absence of humic acid (HA). Two uranyl surface species with fluorescence lifetimes of 5.9 ( 1.4 and 42.5 ( 3.4 µs and 4.4 ( 1.2 and 30.9 ( 7.2 µs were identified in the binary (U(VI)-kaolinite) and ternary system (U(VI)-HA-kaolinite), respectively. The fluorescence spectra of adsorbed uranyl surface species are described with six and five fluorescence emission bands in the binary and ternary system, respectively. The positions of peak maxima are shifted significantly to higher wavelengths compared to the free uranyl ion in perchlorate medium. HA has no influence on positions of the fluorescence emission bands. In the binary system, both surface species can be attributed to adsorbed bidentate mononuclear surface complexes, which differ in the number of water molecules in their coordination environment. In the ternary system, U(VI) prefers direct binding on kaolinite rather than via HA, but it is sorbed as a uranyl-humate complex. Consequently, the hydration shell of the U(VI) surface complexes is displaced with complexed HA, which is simultaneously distributed between kaolinite particles. Aluminol binding sites are assumed to control the sorption of U(VI) onto kaolinite.

Introduction Understanding the migration behavior of radioactive and nonradioactive toxic substances in geological environments is essential for evaluating the safety and reliability of potential long-term nuclear waste disposal sites, of facilities of former uranium mining and milling sites in Saxony and Thuringia (Germany) and of subsurface dumps and sites with radioactive and/or heavy metal inventories. The uranium can also be used as an analogue of actinide ions with oxidation number VI, e.g., Np(VI) or Pu(VI). Clay rocks are considered as a possible host rock for a potential future nuclear waste repository in Germany. Therefore, detailed studies are required on the kinetics, thermodynamics, redox behavior, and speciation of actinides in this rock formation. The migration behavior of actinides in the environment depends on prevailing geochemical conditions and processes influencing the speciation of actinides. One important retardation process is the sorption of actinides onto mineral surfaces. An improved understanding of the mechanisms of radionuclide sorption on clays is based on the identification of the coordination environment of the adsorbed radionuclide species. * Corresponding author phone: +49 351 260 2148; fax: +49 351 260 3553; e-mail address: [email protected]. 6142

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

Interaction processes of actinides are influenced by humic acids (HA, organic macromolecules ubiquitous in natural environments), which can also be a component of clay formations or soils. HA are soluble in the pH range of natural waters and are capable of complexation and colloid formation. Thus, they can affect the transport behavior of metal ions, including radionuclides, due to the formation of soluble complexes with them. Organic compounds, both associated with the clay and released from the clay, can influence the sorption behavior of metal ions such as actinides. Therefore, it is important to study the interaction of metal cations with minerals in ternary systems. An overview of appropriate literature and the results of the U(VI) sorption onto kaolinite (as a model mineral for clays) in the absence and presence of HA studied by batch sorption experiments are given in our previous publication (1). It was shown that the U(VI) sorption onto kaolinite is influenced by the pH, presence of CO2, U(VI) concentration, and the presence of HA. HA enhances the U(VI) uptake in the acidic pH range compared to the system without HA and reduces the U(VI) sorption in the near-neutral pH range. For the modeling of the U(VI) sorption onto kaolinite, information on the structure of the formed surface complexes is necessary. Therefore, in this work TRLFS was used for the first time to characterize the species of U(VI) adsorbed onto kaolinite in the absence and presence of HA. For a better understanding of the sorption mechanism, the obtained results are compared with other model systems representing possible sorption sites of kaolinite. Time-resolved laser-induced fluorescence spectroscopy (TRLFS) is increasingly used to study the speciation of fluorescent actinides or their complexation with organic (e.g., HA, refs 2-5) and inorganic ligands (e.g., sulfate, carbonate, phosphate, refs 6-8). Recently the application of this method was spread to study the character of adsorbed actinide species onto solid matter (9-17). TRLFS was performed to study the U(VI) sorption, for example, onto gibbsite (10, 11), silica (12), and muscovite (13). TRLFS provides information on both lifetime and spectral signature of the adsorbed species, which offers access to a number of different species and their spectral identity (18). Together with EXAFS investigations and other complementary methods (e.g., XPS, ATR-FTIR, NMR), the TRLFS technique can supply new insight into actinide surface complexation, and additionally, it can contribute to an improved knowledge of actinide behavior in the environment. The major advantages of TRLFS over other techniques, such as EXAFS and NMR, are (i) its enhanced sensitivity and (ii) its combined information on concentrations (based on intensities) and coordination (based on emission wave numbers and lifetimes) (12). TRLFS also has the benefit that it is a noninvasive and in-situ method for the direct investigation of solutions, solids, and sorbates. A drawback of this method is the limited number of fluorescent species, thus TRLFS is not a universal method. The measurements are dependent on temperature and strongly influenced by apparatus properties. The derived model and parameters allow a more reliable formulation of surface complexation reactions and consequently lead to an improved modeling of actinide migration in the environment. As a result, more trustworthy safety assessments of potential nuclear waste repositories can be achieved. Furthermore, the results can contribute to the development of efficient remediation measures for radioactively contaminated areas. 10.1021/es070419q CCC: $37.00

 2007 American Chemical Society Published on Web 08/02/2007

Experimental Section Materials. The Georgia kaolinite KGa-1b used in this work was obtained from the Clay Minerals Society Source Clay Repository (IN: 47907-2054). It represents a well-crystallized kaolinite (19). In the experiments, kaolinite was used without any pretreatment. The surface area of kaolinite was determined from the BET (Brunauer-Emmet-Teller) plot of N2 sorption isotherm data. The cation exchange capacity (CEC) was measured by the compulsive exchange method (20). The BET surface area and the CEC of kaolinite were found to be 11.8 m2/g and 1.83 meq/100 g, respectively. Particle measurements indicated that the grain size of KGa-1b amounts to 57.8% < 2 µm and 32.0% < 0.5 µm (19). Synthetic unlabeled (batch M145) HA type M42 (21, 22) and 14C-labeled (batch M170) (22) were applied. Synthetic HA M42 shows an elemental composition similar to natural HA. The carboxyl group content and the proton exchange capacity (PEC) of this synthetic HA type M42 are comparable to those of naturally occurring HA. In addition to that, HA type M42 shows U(VI) complexation behavior comparable to naturally occurring HA (21). For the experiments, a HA stock solution of 5 g/L was prepared by weighing 50 mg HA, adding of 1720 µL of 0.1 M NaOH and filling up the volume with 0.1 M NaClO4 to 10 mL. A 1 × 10-3 M UO2(ClO4)2 (0.01 M HClO4) solution was used as U(VI) stock solution for all the experiments. A 0.1 M NaClO4 (pH ∼ 4.8) solution was prepared by dissolution of NaClO4‚H2O (p.a., Merck, Darmstadt, Germany) in purified Milli-Q-water produced by Milli-RO/Milli-Q-System (Millipore, Schwalbach, Germany). TRLFS: Sample Preparation. TRLFS data were collected for two systems: a binary system, which consists of U(VI) adsorbed on kaolinite, and a ternary system, where HA was also present. U(VI) or U(VI)-HA were adsorbed as described in ref 1. Initial concentrations of U(VI) and HA were 1 × 10-5 M and 10 mg/L, respectively, ionic strength was set to 0.1 M by NaClO4. The pH values were adjusted between pH 5 and pH 8 by addition of appropriate amount of NaOH and/or HClO4 and the solid/solution ratio was 4 g/L. The conditioning time of kaolinite was 72 h and the contact time between U(VI), HA and kaolinite was 60 h. After phase separation (centrifugation for 30 min at ∼2800g and filtration using Minisart N membrane filters (Sartorius, Goettingen, Germany) with a pore size of 450 nm), the supernatants were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) using the model ELAN 6000 (Perkin-Elmer, Boston, MA) for the final U(VI) concentration. The U(VI) sorption onto the vial wall was checked and it reached values up to 5%. For additional details of sample preparation see ref 1. For the spectroscopic investigations, kaolinite samples from the batch experiments were first resuspended in 10 mL of a solution with pH and ionic strength being identical to the original solution (without U(VI) or U(VI)-HA). Timeresolved spectra of this kaolinite suspension were recorded during permanent stirring. The re-distribution of the fluorescent U(VI) species between solid and solution phase was

TABLE 1. Samples for TRLFS Measurements ([U(VI)] ) 1 × 10-5 M, I ) 0.1 M NaClO4, pCO2 ) 10-3.5 atm) sample U5 U6 U7 U8 U-HA5 U-HA6 U-HA7 U-HA8

[HA]0 (mg/L)

pHstart

pHend

U(VI)supernatant (µmol/L)

Usorbed (%)

10 10 10 10

5.05 6.06 7.05 8.08 5.06 6.05 7.04 8.07

5.03 6.03 7.06 8.03 5.01 6.05 7.04 8.01

5.71 1.12 0.72 1.78 0.74 0.19 0.18 2.47

42.3 87.3 91.4 81.4 97.8 98.1 98.2 75.3

negligible. To verify this, kaolinite was separated again from the solution after the TRLFS measurement and the supernatants were measured by ICP-MS for the U(VI) concentration. The results of this analysis confirmed that U(VI) was not desorbed from kaolinite during TRLFS measurements (duration of the measuring of one sample was ∼2 h). To be sure that HA does not desorb from kaolinite during TRLFS measurements in the ternary system, one series of measurements in the ternary system was performed using the 14Clabeled HA type M42. To determine the HA concentration, the supernatants were analyzed by liquid scintillation counting (LSC) using the LSC counter model LS 6000LL (Beckman Coulter, Krefeld, Germany). As for U(VI), the results of the analysis corroborated that HA was not desorbed from kaolinite during TRLFS measurements. The HA sorption onto the vial walls was negligible in the studied pH range. Table 1 shows the experimental conditions of sample preparation, the U(VI) and HA concentrations, pH values and the amount of sorbed U(VI). TRLFS: Measuring System. The TRLFS system consists of a Nd:YAG diode laser (mod. SL 401-20, Spectron Laser Systems, Rugby, United Kingdom) with subsequent fourth harmonic generation. A wavelength of 266 nm (6, 27) was used for the excitation of the samples, providing a maximum signal-to-noise ratio. Time-resolved spectra were recorded with an ICCD-camera (Roper Scientific GmbH, Ottobrunn, Germany) in the wavelength range between 446 and 617 nm with a resolution of 0.168 nm. The delay time after the excitation laser pulse ranges from 0.03 µs to 100.03 µs and the pulse energy was 0.3 mJ. Each spectrum was measured three times. For each spectrum 100 laser shots were averaged. In total, 201 spectra at equidistant delay time steps of 0.5 µs were collected for one time-resolved spectrum. For further details concerning the TRLFS setup, see ref 23.

Results TRLFS measurements provide two kinds of characteristic information: the position of fluorescence emission bands and the fluorescence lifetimes. The positions of fluorescence bands are primary attributes of TRLFS spectrum, whereas the fluorescence lifetime is a secondary feature, because of its dependence on the experimental temperature (24). As

FIGURE 1. TRLFS spectra of U(VI) adsorbed onto kaolinite at pH ∼ 7 in the absence (left) and presence (right) of HA. VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6143

TABLE 2. Fluorescence Lifetimes of the U(VI) Species Sorbed onto Kaolinite in the Absence and Presence of HA, Errors Represent 2σ sample

pH

τ1 (µs)

τ2 (µs)

U5 U6 U7 U8 mean lifetime U-HA5 U-HA6 U-HA7 U-HA8 mean lifetime

5 6 7 8

5.64 ( 0.30 6.88 ( 0.30 5.61 ( 0.24 5.33 ( 0.32 5.87 ( 1.38 3.54 ( 0.98 4.77 ( 0.74 4.54 ( 0.44 4.68 ( 0.90 4.38 ( 1.14

40.80 ( 0.88 41.30 ( 0.94 44.10 ( 1.26 43.70 ( 1.46 42.48 ( 3.34 26.30 ( 0.82 31.40 ( 2.80 32.70 ( 0.90 34.70 ( 5.96 31.30 ( 7.16

5 6 7 8

Kimura et al. (25, 26) determined for the first time for trivalent actinides, the fluorescence lifetime varies depending on the number of neighboring water molecules surrounding the U(VI) atom (27). Such characteristic spectral information is useful for the identification of fluorescent aqueous uranium species, as well as U(VI) surface species adsorbed onto kaolinite. The original TRLFS spectra of the U(VI) sorbed onto kaolinite in the absence and presence of HA at pH 7 are shown in Figure 1 as examples. Obviously, the presence of HA drastically reduces the signal-noise ratio. This effect is typical for the whole pH range investigated within this work. Furthermore, it can be seen that HA quenches the measured fluorescence intensity of the samples. Although the difference in the U(VI) adsorption amounts to 7% (98% vs 91%, see Table 1), the relative fluorescence intensity is almost five times lower in the presence of HA than in its absence (2500 vs 12 000, see Figure 1). This points to differences in the surface speciation induced by HA, and again holds for all pH values from pH 5 to 8. Lifetime Analysis. Based on the TRLFS spectra, the fluorescence decay function was determined. First, a background value was calculated for each sample as the mean value of fluorescence intensities in those wavelength ranges, where no signal was detectable. The calculated background values were fixed during the fitting. The lifetimes of the species were calculated from an exponential decay function, which has the following form for i species:

Y)

∑ A ‚e i

-tx/τi

(1)

n

where Y is the measured fluorescence intensity at the time x, Ai is the fluorescence intensity of the ith species at the time 0, and τi is the fluorescence lifetime of ith species. The lifetime analysis utilized the OriginPro 7.5G software (OriginLab Corp., Northampton, MA). The best approximation for both the absence and presence of HA was a biexponential decay function for all samples indicating the presence of at least two fluorescent species, i.e., one short- (τ1) and one long-lived (τ2). In all cases assuming only a monoexponential decay gave a significantly worse fit. In the binary system, the calculated average fluorescence lifetimes of the short- and long-lived species are τ1 ) 5.87 ( 1.38 µs and τ2 ) 42.48 ( 3.34 µs, respectively, with errors of (throughout the text) 2σ. The lifetime analysis of the ternary system was more problematic due to the noisy spectra. Therefore, it was performed only in a shortened wavelength range: 480-600 nm. The average fluorescence lifetimes amount to τ1 ) 4.38 ( 1.14 µs and τ2 ) 31.30 ( 7.16 µs for the short- and long-lived species, respectively. Results of fluorescence lifetime determinations for all measured samples are summarized in Table 2. In the presence of HA, the obtained fluorescence lifetimes of the long-lived surface 6144

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

FIGURE 2. Fluorescence lifetimes of the short- and long-lived U(VI) surface species in the presence and absence of HA as a function of pH values. Errors represent 2σ.

FIGURE 3. Deconvoluted fluorescence spectra with characteristic positions of the fluorescence emission bands for the sample measured at pH ∼ 7 in the absence of HA. The evaluation was performed with in-house software by Brendler et al. (29). species are not as consistent as in the absence of HA. Moreover, HA decreases the fluorescence lifetimes of both surface species significantly. In Figure 2, the lifetimes of the surface species identified in the presence and absence of HA are depicted as a function of pH. It is evident that the pH values have no significant influence on fluorescence lifetimes of U(VI) long-lived surface species in the binary and also the ternary system. However, there is a slight effect on fluorescence lifetimes of U(VI) shortlived surface species in both of the studied systems. Determination of Fluorescence Emission Bands. Using the internal peak fitting module of the program OriginPro 7.5G, the fluorescence emission bands can be determined. Six fluorescence emission bands were obtained for all measured samples in the absence of HA, which can be described by a set of six peaks in accordance with the work of Bell and Biggers (28). An example of such deconvoluted fluorescence spectra of a single species with their characteristic emission bands is shown in Figure 3 for U(VI) surface species adsorbed onto kaolinite at pH 7 in the absence of HA. As it can be seen, the six observed characteristic fluorescence emission bands are almost identical for both U(VI) surface species. The peak maxima are situated at 486.9 ( 0.9, 501.8 ( 0.6, 520.6 ( 0.9, 541.7 ( 0.7, 567. 8 ( 1.5, and 583.3 ( 0.6 nm, for details, see Table 3. Only the first five peaks will be considered for further discussion because the sixth emission band is very low and broad and therefore uncertain. The positions of the peak maxima are shifted significantly to higher wavelengths relative to the values for the free uranyl ion in perchlorate medium (30). The shifts range from 16.8 nm for the first peak to 9.7 nm for the fifth one. Due to coincidence of all the fluorescence peaks, these two adsorbed U(VI) surface species, i.e., the short-lived and the long-lived, are assumed to be similar with respect to the coordination

TABLE 3. Wavelengths of Peak Maxima Positions (in nm) for Two Fluorescent Components sample

pH

first peak

second peak HA-absence 502.9 501.1 501.8 501.3

521.9 519.0 520.0 519.9

541.5 541.1 541.4 540.8

568.5 570.7 566.8 566.7

583.9 583.9 583.9 583.6

HA-absence 501.3 502.4 501.8 501.9

521.3 521.1 521.0 520.5

542.0 541.5 542.5 542.8

569.1 566.2 567.1 567.3

582.4 582.6 583.4 583.0

U5 U6 U7 U8

5 6 7 8

short-lived species 486.7 486.2 488.7 486.8

U5 U6 U7 U8

5 6 7 8

long-lived species 486.8 486.2 486.5 487.7

third peak

fourth peak

fifth peak

weighted average of all peak maxima with standard deviation 486.9 ( 1.8 501.8 ( 1.2 520.6 ( 1.8 541.7 ( 1.4 567.8 ( 3.0 U-HA5 U-HA6 U-HA7 U-HA8

5 6 7 8

short-lived species 486.4 485.2 488.8 485.9

U-HA5 U-HA6 U-HA7 U-HA8

5 6 7 8

long-lived species 486.7 486.7 487.2 485.7

sixth peak

583.3 ( 1.2

HA-presence 504.9 501.7 500.1 503.1

523.3 519.5 519.5 521.5

540.9 538.6 544.1 542.9

564.0 571.8 566.8 566.9

-

HA-presence 498.7 499.8 501.8 498.9

518.8 521.2 520.1 519.9

546.1 544.4 541.9 539.1

567.1 559.8 570.4 566.9

-

weighted average of all peak maxima with standard deviation 486.6 ( 2.2 501.1 ( 4.4 520.5 ( 3.0 542.2 ( 5.2 566.7 ( 7.2 UO22+

470.1

peak maxima of free UO22+ ion in perchlorate medium (30) 487.8 509.3 532.6

environment throughout the investigated pH range. They should, therefore, have identical numbers of hydroxyl groups in their first coordination sphere, as different numbers of hydroxyl group cause changes in the spectral features (31). In the ternary system, as in the binary system, six peaks were expected, but the position of the peak at the highest wavelength was very uncertain and, thus, only the positions of the first five peaks were determined. Their peak maxima are situated at 486.6 ( 1.1, 501.1 ( 2.2, 520.5 ( 1.5, 542.2 ( 2.6, and 566.7 ( 3.6 nm, see, again, Table 3. No differences in peak positions were observed in comparison to the binary system. It can, therefore, be concluded that HA does not influence the peak positions and the numbers of hydroxyl groups in the first coordination sphere of both species.

Disscusion Two U(VI) surface species with different fluorescence lifetimes were identified by TRLFS measurements for the binary system, U(VI)-kaolinite. The shorter fluorescence lifetime indicates more water molecules in the coordination environment of the respective adsorbed U(VI) surface species, because water molecules quench the fluorescence lifetimes (25, 32). On this basis, it can be concluded that U(VI) forms two surface species on kaolinite that differ in the amount of water molecules in their coordination environment. Furthermore, from EXAFS measurements of U(VI) sorption onto kaolinite (33) it was concluded that U(VI) forms inner-sphere surface complexes. Moreover, two U-Si/Al interactions were found by EXAFS, which means that these surface complexes are bidentate. Thus, U(VI) is associated by edge-sharing with SiO4-tetrahedra and/or AlO6-octahedra. Baumann et al. (10) and Arnold et al. (13) studied the U(VI) sorption on gibbsite ([U(VI)]: 1 × 10-5 M, pH: 5-8.5, I: 0.1 M NaClO4) and muscovite ([U(VI)]: 1 × 10-5 M, pH: 7, I: 0.01 M NaClO4), respectively, by TRLFS. Baumann et al. (10) attributed the surface species with the shorter fluorescence lifetime to the bidentate mononuclear inner-sphere surface complex, in which U(VI) is bound to two reactive hydroxyl groups at the broken edge linked to one Al. Arnold et al. (13) ascribed the

558.1

585.4

surface species with the shorter fluorescence lifetime to the inner-sphere bidentate surface complex, in which U(VI) binds to aluminol groups of edge-surfaces of muscovite. Both interpreted the surface species with significantly longer fluorescence lifetimes as an amorphous U(VI) condensate or nanosized clusters of polynuclear uranyl surface species. This conclusion was made on the basis of findings by Thompson (34) and Sylwester (35). From their EXAFS measurements, performed in the systems U(VI)-kaolinite KGa-2 (34), U(VI)-silica, and U(VI)-γ-alumina (35) they found a U-U interaction at the near-neutral pH range, indicating the formation of polynuclear surface species. This could not be concluded for the second surface species identified in the system studied in this work because in the EXAFS spectra of U(VI) adsorbed onto kaolinite KGa-1b (see ref 33) no U-U interactions were detected. Therefore, U(VI) only forms mononuclear surface complexes on kaolinite. This is supported by the works of Reich (36) and Gabriel (12). Reich et al. (36) measured EXAFS of U(VI) adsorbed onto silica gel. Several samples were measured. One of them was prepared using experimental conditions similar to this work: [U(VI)]: 2 × 10-5 M, pH: 4.5, I: 0.1 M NaClO4, U(VI) sorption was 99%. They did not observe any distinct peak that could be attributed to backscattering from U neighbors. From all their measurements they found that U(VI) forms inner-sphere mononuclear complexes at the silica gel surface. Gabriel et al. (12) studied the uranyl surface speciation on silica particles by means of TRLFS under the following experimental conditions: [U(VI)]: 1 × 10-5 M, 1 × 10-6 M, pH: 4-9, I: 0.01 M NaNO3. Two fluorescent surface complexes were distinguished by their fluorescence lifetimes. However, from the surface complexation modeling of the experimental data they finally identified three surface complexes, postulated as the first fluorescent complex at around pH 5 ≡SiO2UO20, the second fluorescent complex at pH 7.7 ≡SiO2UO2OH-, and the third nonfluorescent uranyl-silica-carbonate complex ≡SiO2UO2OHCO33- occurring around pH 8.6. In the presence of HA, the quenching of fluorescence intensity points to the complexation of U(VI) with HA, thus, VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6145

TABLE 4. Comparison of Mean Values of Fluorescence Lifetimes Obtained for the Systems Studied in This Work with Model Systems UO2-Gibbsite and UO2-Silica Gel system

τ1 (µs)

τ2 (µs)

reference

UO2-gibbsite UO2-silica gel UO2-silica gel UO2-kaolinite UO2-HA-kaolinite

2.0 ( 138.4 ( 105.8b 170 ( 25c 5.9 ( 1.4b 4.4 ( 0.6b

11.1 ( 361.8 ( 206.4b 360 ( 50d 42.5 ( 3.4b 30.9 ( 7.2b

39 40 12 this work this work

a

Average value pH 5 - 7.

0.3a

b

3.8a

Average value pH 5 - 8. c pH 4.41.

d

pH

8.19

U(VI) is adsorbed onto kaolinite as a uranyl-humate complex. Aqueous uranyl-humate complexes themselves do not show any fluorescence at all (36) under the applied measuring conditions, so if only aqueous uranyl-humate complexes would be present, no fluorescence signal would be obtained. There are two possibilities for how U(VI) can be bound to kaolinite in the presence of HA: either it is attached to the kaolinite surface via sorbed HA or it is directly bound to kaolinite with HA additionally attached to uranyl unit. From the EXAFS measurements (23, 32) it was found that even if HA is present, U(VI) prefers to adsorb directly onto kaolinite rather than via HA. This was indicated by the detection of U-Si/Al interactions. Furthermore, it can be proposed that U(VI) forms inner-sphere bidentate mononuclear surface complexes in the ternary as well as in the binary system. Moreover, from XPS (38) measurements it can be concluded that the surface of the kaolinite particles is not covered by a homogeneous HA layer. Part of the HA must be distributed between the kaolinite particles. This means that in the ternary system, U(VI) can interact with significant areas of the kaolinite surface that are not covered by HA. The EXAFS measurements of the U(VI) sorption onto kaolinite in the presence of HA (23) showed that HA has no influence on the U(VI) EXAFS structural parameters in the kaolinite surface complexes. However, in EXAFS it is not possible to distinguish between water molecules and HA coordinated to the uranyl ion primarily because of their very similar values of axial and equatorial U-O distances. The significantly shorter fluorescence lifetimes of both species in comparison to the binary system signals that HA is present in the coordination environment of adsorbed U(VI) surface species, which means that the water molecules of the hydration shell of uranyl ions are partly displaced with HA. To identify which of the possible binding sites of kaolinite U(VI) would prefer to adsorb, TRLFS measurements of the system U(VI)-kaolinite were compared with the systems U(VI)-gibbsite (39) and U(VI)-silica gel (40). Gibbsite, Al(OH)3, was chosen as a model mineral for aluminol sites and silica gel, SiO2, represents silanol binding groups. Silica gel was preferred to quartz due to its high specific area compared to quartz. Table 4 compares the mean values of fluorescence lifetimes of the U(VI)-kaolinite system with these model systems. In all systems at least two fluorescence lifetimes were identified. The fluorescence lifetimes of both fluorescence species on kaolinite lie in the range between values obtained for gibbsite and for silica gel, however, closer to the former ones. It seems that U(VI) adsorbs on both kinds of available sites with the respective ratio being difficult to obtain. From the SEM of kaolinite particles (from our own unpublished work), the ratio between planes and edge areas of kaolinite particles was estimated to be about 0.72, indicating a higher amount of edges relative to basal planes. Brady et al. (41) reported a higher percentage of edges relative to basal planes of the kaolinite KGa-1 using scanning force microscopy. They also reported elevated reactivity of Al edge sites, relative to 6146

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 17, 2007

Si, and the weak sorption on basal planes found from their molecular modeling. Similarly, Zachara and McKinley confirmed in their work the importance of edge coordination reactions for metal ion (especially hydrolyzable metals) adsorption by smectites (42). All this supports the idea that Al binding sites govern the U(VI) sorption onto kaolinite in the system studied in this work. Finally, a stoichiometric excess of Al2O3 in kaolinite was found by chemical analysis (23).

Acknowledgments We thank U. Schaefer for ICP-MS analysis and C. Eckardt for BET measurement. Financial support through the Federal Ministry of Economics and Technology under contract no. 02E9673 is gratefully acknowledged.

Supporting Information Available (A) TRLFS, data processing, fluorescence lifetimes, (B) New TRLFS measurements of U(VI) sorption onto gibbsite, (C) TRLFS properties of U(VI) sorbed onto silica gel. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Krˇepelova´, A.; Sachs, S.; Bernhard, G. Uranium(VI) sorption onto kaolinite in the presence and absence of humic acid. Radiochim. Acta 2006, 94, 825-833. (2) Czerwinski, K. R.; Buckau, G.; Scherbaum, F.; Kim, J. I. Complexation of the uranyl ion with aquatic humic acid. Radiochim. Acta 1994, 65, 111-119. (3) Pompe, S.; Brachmann, A.; Bubner, M.; Geipel, G.; Heise, K.-H.; Bernhard, G.; Nitsche, H. Determination and comparison of uranyl complexation constants with natural and model humic acids. Radiochim. Acta 1998, 82, 89-95. (4) Sachs, S.; Brendler, V.; Geipel, G. Uranium(VI) complexation by humic acid under neutral pH conditions studied by laserinduced fluorescence spectroscopy. Radiochim. Acta 2007, 95, 103-110. (5) Morgenstern, M.; Klenze, R.; Kim, J. I. The formation of mixedhydroxo complexes of Cm(III) and Am(III) with humic acid in the neutral pH range. Radiochim. Acta 2000, 88, 7-16. (6) Geipel, G.; Brachmann, A.; Brendler, V.; Bernhard, G.; Nitsche, H. Uranium(VI) Sulfate complexation studied by time-resolved laser-induced fluorescence spectroscopy (TRLFS). Radiochim. Acta 1996, 75, 199-204. (7) Brendler, V.; Geipel, G.; Bernhard, G.; Nitsche, H. Complexation in the system UO22+/PO43-/OH-(aq): potentiometric and spectroscopic investigations at very low ionic strengths. Radiochim. Acta 1996, 74, 75-80. (8) Bernhard, G.; Geipel, G.; Reich, T.; Brendler, V.; Amayri, S.; Nitsche, H. Uranyl(VI) carbonate complex formation: validation of the Ca2UO2(CO3)3(aq.) species. Radiochim. Acta 2001, 89, 511-518. (9) Froideval, A.; Del Nero, M.; Gaillard, C.; Carillon, R.; Rossini, I.; Hazemann, J. L. Uranyl sorption species at low coverage on Al-hydroxide: TRLFS and XAFS studies. Geochim. Cosmochim. Acta 2006, 70, 5270-5284. (10) Baumann, N.; Brendler, V.; Arnold, T.; Geipel, G.; Bernhard, G. Uranyl sorption onto gibbsite studied by time-resolved laserinduced fluorescence spectroscopy (TRLFS). J. Colloid Interface Sci. 2005, 290, 318-324. (11) Chang, H.-S.; Korshin, G. V.; Wang, Z.; Zachara, J. M. Adsorption of uranyl on gibbsite: a time-resolved laser-induced fluorescence spectroscopy study. Environ. Sci. Technol. 2006, 40, 12441249. (12) Gabriel, U.; Charlet, L.; Schla¨pfer, C. W.; Vial, J. C.; Brachmann, A.; Geipel, G. Uranyl surface speciation on silica particles studied by time-resolved laser-induced fluorescence spectroscopy. J. Colloid Interface Sci. 2001, 239, 358-368. (13) Arnold, T.; Utsunomiya, S.; Geipel, G.; Ewing, R. C.; Baumann, N.; Brendler, V. Adsorbed U(VI) surface species on muscovite identified by laser fluorescence spectroscopy and transmission electron microscopy. Environ. Sci. Technol. 2006, 40, 46464652.

(14) Stumpf, T.; Bauer, A.; Choppin, F.; Fangha¨nel, Th.; Kim, J. I. Inner-sphere, outer-sphere and ternary surface complexes: a TRLFS study of the sorption process of Eu(III) onto smectite and kaolinite. Radiochim. Acta 2002, 90, 345-349. (15) Tits, J.; Stumpf, T.; Rabung, T.; Wieland, E.; Fangha¨nel, Th. Uptake of Cm(III) and Eu(III) by calcium silicate hydrates: a solution chemistry and time-resolved laser fluorescence spectroscopy study. Environ. Sci. Technol. 2003, 37, 3568-3573. (16) Rabung, T.; Stumpf, T.; Geckeis, H.; Klenze, R.; Kim, J. I. Sorption of Am(III) and Eu(III) onto γ-alumina: experiment and modeling. Radiochim. Acta 2000, 88, 711-716. (17) Stumpf, T.; Bauer, A.; Choppin, F.; Kim, J. I. Time-resolved laser fluorescence spectroscopy study of the sorption of Cm(III) onto smectite and kaolinite. Environ. Sci. Technol. 2001, 35, 36913694. (18) Walter, M.; Arnold, T.; Geipel, G.; Scheinost, A.; Bernhard, G.: An EXAFS and TRLFS investigation on uranium(VI) sorption to pristine and leached albite surfaces. J. Colloid Interface Sci. 2005, 282, 293-305. (19) Pruett, R. J.; Webb, H. L. Sampling and analysis of KGa-1b wellcrystallized kaolin source clay. Clays Clay Miner. 1993, 41, 514519. (20) Sumner, M. E.; Miller, W. P. Cation Exchange Capacity and Exchange Coefficients. In Methods of Soil Analysis. Part 3: Chemical Methods; Bigham, J. M., Ed.; Soil Science Society of America: Madison, WI, 1996; pp 1201-1229. (21) Pompe, S.; Brachmann, A.; Bubner, M.; Geipel, G.; Heise, K.-H.; Bernhard, G.; Nitsche, H. Determination and comparison of uranyl complexation constants with natural and model humic acids. Radiochim. Acta 1998, 82, 89-95. (22) Sachs, S.; Schmeide, K.; Brendler, V.; Krˇepelova´, A.; Mibus, J.; Geipel, G.; Heise, K.-H.; Bernhard G. Synthesis and characterization of humic acids with specific functional properties. In Investigation of the Complexation and the Migration of Actinides and Non-Radioactive Substances with Humic Acids under Geogenic Conditions: Complexation of Humic Acids with Actinides in the Oxidation State IV Th, U, Np; Report FZR-399; Forschungszentrum Rossendorf, Institute of Radiochemistry: Dresden, 2004; pp 3-19. (23) Krˇepelova´, A. Influence of Humic Acid on the U(VI) and Am(III) Sorption onto Kaolinite. Ph.D. Thesis, Dresden University of Technology, 2007. (24) Wang, Z.; Zachara, J. M.; Gassman, P. L.; Liu, C.; Catalano, J. Fluorescence spectroscopic studies of uranium-bearing Vadose zone sediments. In Digest of Science and Technology Program Evaluation, Appendix D; Handford Valdoze Zone Report; RPP10098; Pacific Northwest National Laboratory: Richland, WA, 2002. (25) Kimura, T.; Choppin, G. R. Luminescence study on determination of the hydration number of Cm(III). J. Alloys Compd. 1994, 213/214, 313-317. (26) Kimura, T.; Kato, Y. Luminescence study on determination of the inner-sphere hydration number of Am(III) and Nd(II). J. Alloys Compd. 1998, 271 - 273, 867-871. (27) Geipel, G.; Bernhard, G.; Rutsch, M.; Brendler, V.; Nitsche, H. Spectroscopic properties of uranium(VI) minerals studied by time-resolved laser-induced fluorescence spectroscopy (TRLFS). Radiochim. Acta 2000, 88, 757-762. (28) Bell, J. T.; Biggers, R. E. Absorption spectrum of the uranyl ion in perchlorate media. III. Resolution of the ultraviolet band structure; some conclusions concerning the excited state of UO22+. J. Mol. Spectrom. 1968, 25, 312-329. (29) Brendler, V.; Brachmann, A.; Geipel, G. Software to compute fluorescence lifetimes from time-resolved laser-induced fluo-

(30)

(31)

(32)

(33)

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

(42)

rescence spectroscopy (TRLFS) studies, exploiting the full spectrum. In Annual Report 1996; Report FZR-180; Forschungszentrum Rossendorf, Institute of Radiochemistry: Dresden, 1997; p 54. Brachmann, A. Zeitaufgelo¨ste laser-induzierte Fluoreszenzspektroskopie zur Charakterisierung der Wechselwirkung des Uranylions mit Huminsa¨uren und Carboxylatliganden. Ph.D. Thesis, Dresden University of Technology, 1997. Eliet, V.; Bidoglio, G.; Omenetto, N.; Parma, L.; Grenthe, I. Characterization of hydroxide complexes of uranium(VI) by time-resolved fluorescence spectroscopy. J. Chem. Soc. Faraday Trans. 1995, 91, 2275-2285. Bouby, M.; Billard, I.; Bonnenfant, A.; Simonin, J. P.; Bernard, O.; Goez-Grandmont, G. J.; Brunette, J. P. Effects of very high ionic strength on some species of interest in nuclear fuel reprocessing. Czech J. Phys. 1999, 49/S1, 673-678. Reich, T. Ye.; Kokrshunov, M. E.; Antonova, T. V.; Ageev, A. L.; Moll, H.; Reich, T. New regularization method for EXAFS analysis. AIP Conf. Proc. 2007, 882, 179. Thompson, H. A.; Parks, G. A.; Brown, G. E., Jr. Structure and composition of uranium(VI) complexes at the kaolinite-water interface. In Adsorption of Metals by Geomedia: Variables, Mechanisms, and Model Applications; Jenne, E. A., Ed.; Academic Press: San Diego, CA, 1998; pp 349-370. Sylwester, E. R.; Hudson, E. A.; Allen, P. G. The structure of uranium (VI) sorption complexes on silica, alumina, and montmorillonite. Geochim. Cosmochim. Acta 2000, 64, 24312438. Reich, T.; Moll, H.; Arnold, T.; Denecke, M. A.; Hennig, C.; Geipel, G.; Bernhard, G.; Nitsche, H.; Allen, P. G.; Bucher, J. J.; Edelstein, N. M.; Shuh, D. K. An EXAFS study of uranium(VI) sorption onto silica gel and ferrihydrite. J. Electron Spectrosc. Relat. Phenom. 1998, 96, 237-243. Sachs, S.; Brendler, V.; Geipel, G. Uranium(VI) complexation by humic acid under neutral pH conditions studied by laserinduced fluorescence spectroscopy. Radiochim. Acta 2007, 95, 103-110. Reich, T.; Drebert, J.; Krˇepelova´, A.; Sachs, S. XPS Study of humic acid/kaolinite interaction. In Annual Report 2005; Report FZR443; Forschungszentrum Rossendorf, Institute of Radiochemistry: Dresden, 2006; p 46. Krˇepelova´, A.; Brendler, V.; Sachs, S.; Baumann, N.; Bernhard, G. New TRLFS measurements of U(VI) sorption onto gibbsite. In Annual Report 2006; Report FZD-459; Forschungszentrum Dresden-Rossendorf, Institute of Radiochemistry: Dresden, 2007; p 54. Trepte, P. Sorption von Radionukliden an Tongestein: Spektroskopische Referenzdaten. Diploma Thesis, University of Applied Science, Dresden, 2006. Brady, P. V.; Cygan, R. T.; Nagy, K. L. Surface charge and metal sorption to kaolinite. In Adsorption of Metals by Geomedia: Variables, Mechanisms, and Model Applications; Jenne, E. A., Ed.; Academic Press: San Diego, CA, 1998; pp 371-382. Zachara, J. M.; McKinley, J. P. Influence of hydrolysis on the sorption of metal cations by smectites: importance of edge coordination reactions. Aquat. Sci. 1993, 55, 250-261.

Received for review February 19, 2007. Revised manuscript received June 8, 2007. Accepted June 20, 2007. ES070419Q

VOL. 41, NO. 17, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6147