Humic Colloid-Borne Natural Polyvalent Metal Ions ... - ACS Publications

Jump to Experimental Section - Manual mixing of the suspension is done occasionally every 2 d. All experiments are carried out at room temperature (25...
1 downloads 0 Views 153KB Size
Environ. Sci. Technol. 2002, 36, 2946-2952

Humic Colloid-Borne Natural Polyvalent Metal Ions: Dissociation Experiment H . G E C K E I S , * ,† T H . R A B U N G , † T. NGO MANH,‡ J. I. KIM,† AND H. P. BECK‡ Forschungszentrum Karlsruhe GmbH, Institut fu ¨ r Nukleare Entsorgung, P.O. Box 3640, D-76021 Karlsruhe, Germany, and Universita¨t des Saarlandes, Institut fu ¨ r Anorganische und Analytische und Radiochemie, D-66123 Saarbru ¨ cken, Germany

The natural association nature of the humic colloid-borne trace elements is investigated. Rare earth elements (REE) Th and U are chosen as naturally occurring representatives and chemical homologues for actinides of different oxidation states present in nuclear waste. Tri- and tetravalent elements in two investigated Gorleben groundwaters (Gohy532 and -2227) almost exclusively occur as humic or fulvic colloid-borne species. Their desorption behavior from colloids is examined in the unperturbed groundwater (pH ∼8) under anaerobic conditions (Ar/1% CO2) by addition of a chelating cation exchanger resin. Particularly, the dissociation process of naturally occurring Eu(III) in the groundwater is compared with the Eu(III) desorption from its humate complex prepared with purified Aldrich humic acid in a buffered aqueous solution at pH ∼8. The Eu(III) dissociation from the groundwater colloids is found to be considerably slower than found for the humate complex synthesized in the laboratory. This suggests that under natural aquatic conditions the Eu(III) binding in colloids is chemically different from the simple humate complexation as observed in the laboratory experiment. The colloid characterization by the size exclusion chromatography (SEC) and the flow field-flow fractionation (FFFF) indicates that natural colloid-borne trace elements are found predominantly in colloids of larger size (>15 nm in size), while Eu(III) in its humate complex is found mainly in colloids of hydrodynamic diameters 83%) and minimal charge repulsion effects. Details of the method are described

elsewhere (15). A symmetric fractionator is used (FFFractionation, Inc., Salt Lake City, UT). A regenerated cellulose membrane with a nominal cutoff (CO) of 5 kDa from Schleicher & Schuell (Dassel, Germany) represents the accumulation wall. The carrier consists of a 0.005 mol/L Tris buffer (pH 9) solution and is degassed by a 1100 series Vacuum Degasser model G 1322A. The carrier flow is delivered at a constant rate to the fractionator by a 1100 HPLC Iso pump model G 1310A from Hewlett-Packard (Waldbronn, Germany). From the channel, the effluent is directed through an UV/VIS detector (Lambda-Max LC model 481, Waters, Milford, MA) to determine humic and fulvic acids at λ ) 254 nm. The cross-flow is provided by a double piston precision pump P-500 from Pharmacia Biotech AB (Sweden) in a recirculating cross-flow loop. Potential contamination of the cross-flow cycle by sample components passing the ultrafiltration membrane was not observed by temporary UV/VIS and ICP-MS measurements. Channel-flow is kept fixed at 1 mL/min, and the cross-flow rate is adjusted to 5 mL/min. To elute larger particles, the cross-flow is decreased after 15 min to 0.5 mL/min. A total of 100 µL of the sample solution is injected for each analysis. The actual channel thickness of w ) 181 ( 0.9 µm is determined experimentally by injecting NaN3 (20 µL, 0.01 w/v % in ultrapure water). The colloid size can be calculated from the elution volume obtained for the respective colloid species (T ) 298.15 K, η ) 0.0089 g cm-1 s-1) according to

VR )

νcw2 πηw2νc dh ) 6D kT 2

(1)

where η is the medium viscosity, νc is the volumetric crossflow, dh is the hydrodynamic diameter of the colloidal species, VR is the retention volume, k is the Boltzmann constant, T is the absolute temperature, and D is the diffusion coefficient. According to eq 1, larger size colloids are eluted later than smaller ones. For the size distribution characterization by SEC, a polymer-based TSK-30 XL column (TosoHaas, Stuttgart, Germany) is used with a nominal CO of 800 kDa related to globular proteins. The carrier solution is a 0.005 mol/L Tris buffer solution (pH 9.1) in which the ionic strength has been slightly increased by addition of 0.01 mol/L NaClO4 in order to avoid ionic exclusion effects. The flow rate is fixed to 1 mL/min. The sequence of elution is different from that in FFFF. Because of the size exclusion mechanism, larger size particles are eluted prior to smaller size colloids. Before injection to either SEC or FFFF, samples are filtered through a filter of 200 nm pore size. To determine the inorganic element content of individual colloid size fractions generated by both fractionation methods, an ICP-MS (Elan 6000, Perkin-Elmer) is directly coupled to the effluent from the UV/VIS detector. A 5% ultrapure nitric acid solution containing 50 µg/L Rh as an internal standard is mixed to the FFFF effluent via a T-piece. A flow of acid solution is provided by a peristaltic pump at a constant rate of 0.5 mL/min. Acidification of the effluent is required to prevent memory effects due to sorption of metal ions to the inner surfaces of the nebulizer and spray chamber of the ICP-MS cross-flow sample introduction system as described in refs 16 and 17. The Rh signal can be used to correct for drifts in the ICP-MS signal during the fractionation (16).

Speciation Calculations The MINEQL code is applied for the speciation calculation (18). An average number of 5.5 mmol/g proton exchanging groups acting as complexing ligands is taken for humic acid (19). The Eu(III) complexation to humic acid is calculated according to the metal ion charge neutralization model (20). VOL. 36, NO. 13, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2947

TABLE 2. Thermodynamic Constants Used for Speciation Calculationsa reaction

log K

ref

H2O S H2CO3 S 2H+ + CO32Eu3+ + CO32- + HA S [Eu(CO3)HA] Eu3+ + 2H2O + HA S [Eu(OH)2HA] + 2H+ Eu3+ + H2O + HA S [Eu(OH)HA] + H+ Eu3+ + HA S [EuHA] Eu3+ + 3CO32- S [Eu(CO3)3]3Eu3+ + 2CO32- S [Eu(CO3)2]Eu3+ + CO32- S [Eu(CO3)]+ Eu3+ + 2H2O S [Eu(OH)2]+ + 2H+ Eu3+ + H2O S [Eu(OH)]2+ + H+ H2IDA S IDA2- + 2H+ HIDA- S IDA2- + H+ Eu3+ + IDA2- S [Eu(IDA)]+ Eu3+ + 2IDA2- S [Eu(IDA)2]Eu3+ + 3IDA2- S [Eu(IDA)3]3-

-13.79 -17.43 12.4 -10.03 -0.96 6.4 13.53 10.81 6.38 -16.15 -8.08 -11.95 -9.34 6.73 12.11 15.79

18 18 39 39 39 20 40 40 40 40 40 21 21 21 21 21

OH- +

a

H+

FIGURE 1. Fractions of Eu(III) species in solution as a function of the contact time with Chelex resin and the equilibration time of the Eu(III) humate complex ([HA] ) 30 mg/L; [Eu(III)] ) 1 × 10-6 mol/L; pH 8.0; I ) 0.1 M NaClO4). For comparison, the Eu(III) sorption onto Chelex resin in the absence of humic acid is shown.

H2IDA, iminodiacetic acid; HA, humic acid (ionic strength: 0.1 mol/L).

Chemical reactions and their constants are given in Table 2. Complexation constants are calculated for ionic strength of 0.1 mol/L using the Davis equation. The constants for the formation of Am(III)/Cm(III) humate and ternary carbonatoand hydroxohumate complexes are found to be very similar for experiments made in solutions of 0.02 and 0.1 mol/L (12, 39). Therefore, the influence of the different ionic strength in the “laboratory” system and the groundwater on the complexation constants is considered negligible. An equilibrium with the respective atmospheric CO2 partial pressures (0.03% for the laboratory system and 1% for the groundwater) is assumed. The interaction of Eu(III) with Chelex resin is treated as complexation with IDA, and data are taken from ref 21. Even though this appears to be a quite simplified view, it is supported in the literature that sorption of trivalent metal ions to IDA resins correlates fairly well with the complexation with IDA in solution (22).

Results and Discussion Laboratory System. Neither sorption of humic acid onto the resin nor degradation of the resin is found during the experiment as confirmed by monitoring the DOC concentration in solution. The DOC concentration of humic colloidcontaining solutions in contact with Chelex resin remains virtually constant during the experiment. Contacting the purified resin with pure water does not yield a DOC increase. In all experiments pH is maintained at around 8 to make comparable with the natural groundwater. A relatively fast reaction is found for the sorption of Eu(III) to the Chelex resin in the absence of humic acid (Figure 1). The Eu(III) concentration decreases rapidly to below 5% of the original concentration and after 3 d below the detection limit of ICPMS. Results from the Aldrich HA/Eu(III) system are included in Figure 1. The sorption of Eu(III) to the resin is delayed in the presence of humic acid depending on the equilibration time of Eu(III) with humic acid prior to the contact with the Chelex resin. In all experiments, a similar Eu(III) concentration is attained after ca. 1000 h (Figure 2). As the sorption reaction of Eu(III) with the chelating resin in absence of humic acid according to

Mx+ + Na/H - Chelex f M - Chelex + xNa+/H+ (2) is rather fast, the dissociation reaction of Eu(III) from its humate complex is observed in the experiment. As a result, it appears that with increasing aging time of the Eu(III) humate complex, the Eu(III) dissociation rate slows down. 2948

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002

FIGURE 2. Fractions of Eu(III) species in humic colloid containing solutions as a function of the contact time with Chelex resin. Experiments with Aldrich HA (laboratory system) and the natural Gorleben groundwaters Gohy-532 and -2227 (natural system) are compared. The desorption of Eu-155 spiked to Gohy-2227 is compared with the behavior of natural Eu(III). To quantify the differences in the metal ion dissociation behavior in the laboratory and natural systems, the dissociation rate of Eu(III) from its humate complex is described by a pseudo-first-order kinetics (11). At least two different Eu(III) humate species, showing “fast” and “slow” dissociation kinetics, are necessary to fit the experimental results:

( )

( )

csol(t) ceq -t -t ) + A1 exp + A2 exp ctot ctot τ1 τ1

(3)

where csol(t) is the concentration of Eu(III) in solution (colloidal and ionic species) at time t (h), ceq is the final concentration of Eu(III) in solution after equilibration with the ion exchanger resin, ctot is the initial concentration of Eu(III) in solution (t ) 0), A1 is the fraction of Eu(III) humate dissociating with a time constant τ1 (%), A2 is the fraction of Eu(III) humate dissociating with a time constant τ2 (%), and τ1 and τ2 are dissociation time constants (h). The results obtained by fitting the experimental data to the kinetic model (eq 3) are listed in Table 3. The experimental data suggest that a steady state for the dissolved Eu(III) species concentration is established at about 13% of the originally present Eu(III) ()ceq). For the discussion of the experimental findings, it appears necessary to inspect the chemical speciation of the metal ion. A quite complete set of thermodynamic data exists for trivalent actinides and Eu(III) (see Table 2), so that a speciation calculation can be made. A mixture of the ternary carbonatohumate and hydroxohumate complexes prevail according to the calculation (Figure 5A). Calculation con-

TABLE 3. Kinetic Parameters Obtained for the Laboratory System by Fitting Experimental Data to the Kinetic Equation (eq 4)a

a

system

τ1 (h)

τ2 (h)

A2

Eu(III) humate; eq time ) 1 h Eu(III) humate; eq time ) 24 h Eu(III) humate; eq time ) 168 h Eu(III) humate; eq time ) 1008 h

4.6 ( 0.5

49 ( 2

A1

139 ( 19

32.6 ( 2

ceq (%)

4.6 ( 0.4

41.7 ( 2

141 ( 10

42 ( 1

13.1 ( 0.8

4.7 ( 0.6

33 ( 2

115 ( 9

50.7 ( 2

12.5 ( 0.8

3.6 ( 1

25 ( 3

150 ( 25

12.8 ( 1

52 ( 3

19 ( 2

Calculations were made using the Fit Routine of ORIGIN Version 4.1. Errors correspond to the uncertainty of fit calculation.

FIGURE 3. Fractions of fast (A1) and slow (A2) dissociating Eu(III) as a function of aging time of Eu(III) humate calculated according to eq 4.

FIGURE 5. Relative species distribution of Eu(III) in the laboratory system in absence (A) and presence (B) of Chelex resin. The gray shaded area marks the pH region of experiment.

FIGURE 4. Fractograms of the Eu(III) humate in the laboratory system after different equilibration times: (A) fractograms detected by the UV/VIS detector correspond to the size distribution of the humic acid; (B) fractogams detected by ICP-MS corresponding to the size distribution of the Eu(III) humate species. The gray shaded areas indicate regions where peak maxima for humic acid and the Eu(III) humate species are situated. For further explanations, see text. sidering the presence of the chelating resin (Figure 5B) suggests the quantitative sorption of Eu(III) onto the resin under given conditions, which is not in total agreement with the experiment, where about 13% remain still dissolved. For the description of the time dependent desorption from humic colloids, the measured Eu(III) concentration remaining in solution is assumed to correspond to an equilibrium concentration (ceq). The decrease of the Eu(III) concentration is then described by using the same time constants (τ1) for the fast and (τ2) for the slow dissociating component irrespective of different aging times. Taking all these findings into account, the data imply that Eu(III) moves from a fast (A1) to a slow (A2) dissociating species with increasing the

reaction time, i.e., the availability of the colloid-borne Eu(III) toward sorption to the Chelex resin decreases with aging time (Figure 3). Findings from kinetic investigations by other authors (5) are consistent with those of the present study. In the Am(III) migration experiment with the Gorleben groundwaters Gohy-532 and -2227 in sandy sediment columns, dissociation time constants in the same order of magnitude (τ1 ) 3 h, τ2 ) 250 h) are required for the parametrization of the kinetic model as found in this work for the laboratory system. Somewhat higher values (666 h) for the slow dissociation rate are reported by King et al. (23) by using a similar experimental approach. The aging effect of Am(III) humate observed in ref 5 on the dissociation behavior is also found to be qualitatively the same. Both experiments, the column experiments with natural groundwater and natural sand as well as the batch experiments in our laboratory system, show consistent kinetic effects as previously reported by different authors (6-11). The variation of colloid sizes in the laboratory system is studied as a function of the equilibration time of Eu(III) with humic acid by FFFF with on-line UV and ICP-MS detection. During the experimental time period, only a slight change of the humic acid size distribution is observed. A shoulder appears at higher elution volume, indicating the generation of some humic acid agglomerates upon addition of Eu(III) (Figure 4A). However, a significant shift of the Eu(III) signal VOL. 36, NO. 13, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2949

FIGURE 6. Fractions of dissolved and colloidal Th, Eu, U, Pb, and Sr species as a function of the contact time with Chelex resin in the groundwater Gohy-2227 (pH 8.1; Ar/1% CO2 atmosphere). Analytical uncertainties are indicated by the error bars. to higher elution volume becomes visible from the respective ICP-MS fractograms (Figure 4B). According to the FFFF equation (eq 1), the mean hydrodynamic diameter of humic acid appears slightly changed during the experiment from 2.3 to 2.6 nm. The Eu(III)-bearing humic acid fraction lies close to that of humic acid, but its size maximum is shifted to larger colloids with increasing the aging time. The mean diameter (size at peak maximum) of humic colloids containing Eu(III) changes significantly from 3.2 up to 3.8 nm. To minimize the influence of drifts and variations in the FFFF conditions on the fractograms, the different samples are measured at the same day in one sequence. The statistical uncertainty for the position of the sample peak maxima determined in triplicate FFFF measurements lies at 17 nm are eluted from the channel. Analysis by FFFF under constant cross-flow conditions gave a size of these particles in the range of 35 nm (29). In this area only a very small concentration of humic colloids is detected by the UV/VIS detector, close to the detector baseline. From the direct comparison with the fractogram of the laboratory system (Figure 8A), the polyvalent metal species in the natural groundwater are located in much larger colloid size fractions than observed in the laboratory system. The FFFF study suffers somewhat from the high dilution of the sample (about 1:500) during the fractionation and the relatively high background concentration of Al and Si due to corrosion of ceramic components of the FFFF channel (28). Therefore, an additional study is made by using SEC, where the dilution of the sample is only about 1:20. Figure 9 shows that the polyvalent metal ions as Th, REE, partially Al, and to a lesser extent Fe eluted earlier than the peak maximum observed for humic acid. Th, REE, and partly Al are eluted close to or even before the nominal exclusion limit of the column. These elements are therefore attached to colloidal species larger than 800 kDa corresponding to a hydrodynamic diameter of >16 nm. In another SEC study (29), it has been shown that under the same experimental conditions as applied in the present study Th-228 and Eu-155 spiked to an aliquot of the Gohy-2227 groundwater bound to smaller colloid sizes as compared to the naturally present elements. The spiked isotopes remain adsorbed to small humic colloids as visible by the UV/VIS detection even after an aging time of 200 d. Consistent to the observation made in the present desorption experiments, the isotopic exchange does not take place as evidenced by the SEC experiment. The size distribution observed for the U species by SEC is distinctly different from that observed for REE and Th. Only a part of U follows the behavior of REE and Th, while another part appears distributed in the range of the humic colloid peak. It can be explained by different oxidation states of U possibly coexisting in groundwater (see discussion

FIGURE 9. Size exclusion chromatograms of Gohy-2227 as recorded by an UV/VIS detector and ICP-MS. The working range of the column and the corresponding hydrodynamic diameters are indicated by the dotted lines. The SEC of the Th-228 and Eu-155 spiked sample is taken from ref 29. above), where U(IV) might be associated with the larger colloidssconsistent to the Th(IV) behaviorsand U(VI) in the smaller sized humic or even ionic fraction. Both Al and U elution bands range up to a total pore volume of the column, i.e., the size of Al and U species is distributed down to that of ionic species. Elution of Si can only be observed at rather high elution volume (Figure 9), indicating the presence of mainly noncolloidal orthosilicilic acid. The early elution of REE, Th, and a part of the U together with Al indicates the presence of inorganic colloids hosting natural actinides and actinide homologues. Because of the relatively high Si background in the SEC eluent, the possibly existing Si containing colloidal species are not detected by the analytical techniques applied and has to be ascertained in further studies. Influence of Colloid Size on Dissociation Kinetics. The increasing kinetic stability of the metal ion humate complexes with increasing equilibration time has been interpreted as a ripening process, where after a rapid initial sorption step, the metal ion either migrates to stronger binding sites (30) or moves to inner sites of the macromolecular structure of the humic colloid (6-11). At the given experimental conditions of rather high pH, the humic acid molecule should have a rather open structural configuration similar to an unfolded coil. As aquatic fulvic/humic acids show a size in the range of clearly less than 10 nm (15, 31), it is difficult to assume the presence of an interior humic area. Intra- and intermolecular bridging by polyvalent cations leading to coiling or agglomeration would be more plausible. Such processes have been reported recently (32) and are even assumed to be relevant for the regulation of the humic matter cycle in the marine environment (33, 34). The presence of a range of chemically different binding sites in humic colloids has also been discussed by various authors (30, 35). Spectroscopic studies over a wide concentration range, however, could not prove their existence for trivalent actinide and Eu(III) ions (36). Only the process of cation-induced agglomeration is able to consistently explain the shift of the Eu humate species toward larger colloid sizes as observed in the FFFF study (Figure 4). The inclusion of the polyvalent metal ion inside humic colloid agglomerates may then be VOL. 36, NO. 13, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2951

responsible for the kinetic hindrance of the Eu(III) dissociation in the laboratory system.

(15) Ngo Manh, Th.; Geckeis, H.; Kim, J. I.; Beck, H. P. Colloids Surf. A 2001, 181, 289-301.

The significantly reduced dissociation rate of Eu(III) from the groundwater humic colloids shows clearly that the chemical state of Eu(III) in the laboratory system is different from that in the natural system. Even though the size fractionation process is very different for FFFF and SEC, both methods show that the polyvalent cations are mainly located in larger colloids in the natural system. The correlation of different element peak positions in the SEC-ICP-MS chromatograms suggests the association of REE, Th, and a part of U to inorganic colloids, consisting of some Al (Figure 9). Inorganic platelet particles with humic colloids arranged at the edges have indeed been identified at the nanoscale in Gohy-2227 groundwater by a recent atomic force microscopy study (37). The humic acid is found to stabilize inorganic colloids as examined also for aquatic colloids in seapage water by transmission electron microscopy (38). The very slow dissociation of naturally abundant metal ions from aquatic humic colloids can be attributed to their inclusion into humic acid stabilized inorganic colloidal species.

(16) Hasselo¨v, M.; Lyven, B.; Haraldsson C.; Sirinawin, W. Anal. Chem. 1999, 71, 3497-3502.

The presence of colloid-borne actinide ions, where metal binding to aquatic colloids is considerably kinetically stabilized or even irreversible, is considered as one of the key uncertainties in the safety assessment of nuclear waste disposal. Further investigations have to concentrate on the quantification of such processes.

Acknowledgments The support of the Adenauer Foundation by granting a Ph.D. stipend for Th.N.M. is gratefully acknowledged.

Note Added after ASAP

(17) Heumann, K. G.; Rottmann, L.; Vogl, J. J. Anal. At. Spectrom. 1994, 9, 1351-1355. (18) Schecher, W. D.; McAvoy, D. C. MINEQL+, User’s Manual; Environmental Research Software: Hallowell, ME, 1994. (19) Kim, J. I.; Buckau, G.; Li, G. H.; Duschner, H.; Psarros, N. Fresenius J. Anal. Chem. 1990, 338, 245-252. (20) Czerwinski, K. R.; Kim, J. I.; Rhee, D. S.; Buckau, G. Radiochim. Acta 1996, 72, 179. (21) Martell, A. E.; Smith, R. M. NIST Critically Selected Stability Constants of Metal Complexes Database, Version 2.0; U.S. Department of Commerce: Gaithersburg, MD, October 1995. (22) Yuchi, A.; Sato, T.; Morimoto, Y.; Mizuno, H.; Wada, H. Anal. Chem. 1997, 69, 2941-2944. (23) King, S. J.; Warwick, P.; Hall, A.; Bryan, N. D. Phys. Chem. Chem. Phys. 2001, 3, 2080-2085. (24) Zeh, P.; Czerwinski, K. R.; Kim, J. I. Radiochim. Acta 1997, 76, 37-44. (25) Buckau, G., Ed. Effects of humic substances on the migration of radionuclides: complexation and transport of actinides; European Commission Project Report EUR 19610 EN; 2000. (26) Morgenstern, A. Humat- und Phosphatkomplexierung von Actinidionen im grundwasserrelevanten pH-Bereich. Ph.D. Thesis, Technische Universita¨t Mu ¨ nchen, 1997. (27) Artinger, R.; Kienzler, B.; Schu ¨ ssler, W.; Kim, J. I. J. Contam. Hydrol. 1998, 35, 261-275. (28) Ngo Manh, T.; Knopp, R.; Geckeis, H.; Kim, J. I.; Beck, H. P. Anal. Chem. 2000, 72, 1-5.

On the fifth page in the next to the last paragraph, Sr2- was changed to Sr2+. The correct version was posted on June 3, 2002.

(29) Bouby, M.; Ngo Manh, T.; Geckeis, H.; Scherbaum, F.; Kim, J. I. Radiochim. Acta (in press).

Literature Cited

(31) Lead, J. R.; Wilkinson, K. J.; Balnois, E.; Cutak, B. J.; Larive, C. K.; Assemi, S.; Beckett, R. Environ. Sci. Technol. 2000, 34, 13651369.

(1) Kim, J. I. Mater. Res. Soc. Symp. Proc. 1993, 294, 3-21. (2) Kim, J. I. MRS Bull. XIX 1994, 12, 47-53. (3) Degueldre, C.; Triay, I.; Kim, J. I.; Vilks, P.; Laaksoharju, M.; Miekeley, N. Appl. Geochem. 2000, 15, 1043-1051. (4) Kim, J. I.; Zeh, P.; Delakowitz, B. Radiochim. Acta 1992, 58/59, 147-154. (5) Schu ¨ ssler, W.; Artinger, R.; Kienzler, B.; Kim, J. I. Environ. Sci. Technol. 2000, 34, 2608-2611. (6) Cacheris, W. P.; Choppin, G. R. Radiochim. Acta 1987, 42, 185190. (7) Choppin, G. R.; Nash, J. Inorg. Nucl. Chem. 1981, 43, 357-359. (8) Choppin, G. R. Radiochim. Acta 1988, 44/45, 23-28. (9) Clark, S. B.; Choppin, G. R. Humic and Fulvic Acids; ACS Symposium Series 651; American Chemical Society: Washington, DC, 1996; pp 207-219. (10) Choppin, G. R.; Clark, S. B. Mar. Chem. 1991, 36, 27-38. (11) Rao, L.; Choppin, G. R.; Clark, S. B. Radiochim. Acta 1994, 66/67, 141-147. (12) Morgenstern, A.; Klenze, R.; Kim, J. I. Radiochim. Acta 2000, 88, 7-16. (13) Rabung, Th. Einfluss von Huminstoffen auf die Europium(III)Sorption an Ha¨matit (Humate and phosphate complexation of actinide ions in a pH range relevant to natural groundwater). Ph.D. Thesis, Universita¨t Saarbru ¨ cken, 1998. (14) Artinger, R.; Kienzler, B.; Schu ¨ ssler, W.; Kim, J. I. In Effects of Humic Substances on the Migration of Radionuclides: Complexation and Transport of Actinides; Buckau, G., Ed.; First Technical Progress Report; FZKA 6124; Forschungszentrum Karlsruhe: August 1998; pp 23-43. 2952

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002

(30) Rate, W.; McLaren, R. G.; Swift, R. S. Environ. Sci. Technol. 1993, 27, 1408-1414.

(32) Engebretson, R.; Von Wandruszka, R. Environ. Sci. Technol. 1998, 32, 488-493. (33) Chin, W.-C.; Orellana, M. V.; Verdugo, P. Nature 1998, 391, 568572. (34) Wells, M. L. Nature 1998, 391, 530-531. (35) Langford, C. H.; Gutzman, D. W. Anal. Chim. Acta 1992, 256, 183-201. (36) Kim, J. I.; Rhee, D. S.; Wimmer, H.; Buckau, G.; Klenze, R. Radiochim. Acta 1993, 62, 35-43. (37) Plaschke, M.; Ro¨mer, J.; Kim, J. I. Environ. Sci. Technol. (submitted for publication). (38) Wilkinson, K. J.; Negre, J. C.; Buffle, J. J. Contam. Hydrol. 1997, 26, 229-243. (39) Panak, P.; Klenze, R.; Kim, J. I. Radiochim. Acta 1996, 74, 141146. (40) Neck, V.; Fangha¨nel, Th.; Kim, J. I. Aquatische Chemie und thermodynamische Modellierung von trivalenten Actiniden (Aquatic chemistry and thermodynamic modeling of trivalent actinide ions); Wissenschaftliche Berichte, FZKA 6110; Forschungszentrum Karlsruhe: 1998.

Received for review December 19, 2001. Revised manuscript received April 22, 2002. Accepted April 29, 2002. ES010326N