Environ. Sci. Technol. 2007, 41, 478-485
Diffusion of 22Na and 85Sr in Montmorillonite: Evidence of Interlayer Diffusion Being the Dominant Pathway at High Compaction MARTIN A. GLAUS,* BART BAEYENS, MICHAEL H. BRADBURY, ANDREAS JAKOB, LUC R. VAN LOON, AND ANDRIY YAROSHCHUK Laboratory for Waste Management, Paul Scherrer Institut, CH-5232 Villigen, Switzerland
A mechanistic understanding of transport phenomena in compacted clays is essential for the use of such materials as engineered barrier systems for the safe geological disposal of radioactive wastes. The present contribution is a first step in the development of an integrative treatment of the properties of tracer cations in compacted bentonites with respect to diffusion and sorption. The diffusion of 22Na and 85Sr in highly compacted montmorillonite and kaolinite is investigated as a function of the “external salt concentration” (NaClO4), i.e., of the solution in equilibrium with the clay. Consistent results were obtained from throughdiffusion experiments and tracer profile analysis. Knowledge of genuine diffusion coefficients of the filter plates turned out to be crucial in cases where the diffusive resistance of the filter plates was similar to that of the clay. Diffusion coefficients formally calculated on the basis of the tracer concentration gradient in the external aqueous phase, and the sorption distribution ratios were found to decrease with increasing external salt concentration in the case of montmorillonite. In a logarithmic representation of these data, a slope of -1 was obtained for the monovalent 22Na, whereas the slope was -2 for the divalent 85Sr. In the case of kaolinite, diffusion coefficients were independent of the external salt concentration. It is postulated that the diffusion of the tracer cation through the interlayer water is the dominant pathway in compacted swelling clays under the experimental conditions tested. Effective diffusion coefficients, based on a tracer concentration gradient in the interlayer water of the clay, were found to be independent of the composition of the external aqueous phase. The latter gradient is assumed to be a function of the external salt concentration, according to a calculated distribution of the tracer cation between free pore water and the interlayer water via cation exchange.
Introduction Compacted bentonites are widely proposed as effective technical barriers retarding radionuclide migration out of * Corresponding author tel.: +41-56-310 22 93; fax: +41-56-310 35 65; e-mail:
[email protected]. 478
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007
the near-field of a deep geological repository. Understanding the migration processes of radionuclides in compacted bentonites is therefore an important requirement for a quantitative assessment of the overall performance of such a repository. The clay fraction of bentonites is composed essentially of montmorillonite, a smectite-type clay mineral, which is composed of layers of octahedrally coordinated aluminum sandwiched by two layers of tetrahedrally coordinated silicium atoms in a 2:1 arrangement (often referred to as TOT sheets). An intrinsic property of these clays is their ability to swell in the presence of water, which leads to compacted bentonites having extremely low hydraulic conductivities (1). Consequently diffusion is assumed to be the dominant migration process of radionuclides across compacted bentonites. Despite the fact that diffusion data for many radionuclides are available in the literature, knowledge on the diffusion process of charged species through compacted clays is still rather poor on the basis of process understanding (2). The diffusion of charged species in compacted bentonites compared to diffusion in free water is influenced by geometrical factors caused by the complex microstructure of these clays, which is characterized by a dense house-of-cards arrangement of clay platelets. These platelets are composed of stacks of TOT sheets. Owing to isomorphic substitution of aluminum and silicium atoms by cations with lower charges, these minerals contain a substantial excess of fixed negative charges that have to be compensated by dissolved countercations. Different physicochemical states of water can be distinguished in smectitic clays. Water may be present as “free water” in the pore space between the clay platelets and is assumed to have properties similar to bulk water. In contrast, water present between the TOT layers, the so-called interlayer water, is influenced by the large amount of chargecompensating cations. The physicochemical properties of interlayer water may distinctly differ from bulk water. A comprehensive analysis of literature data (3) for the diffusion of water molecules in compacted bentonites has shown that these clays can be treated conceptually as a double-porous medium, where water molecules may diffuse in macropores containing free water and in nanopores formed by the interlayer water. As a further consequence of the high density of fixed negative charge, only neutral and cationic species are supposed to diffuse in both the interlayer water and the free pore water. Anions are excluded from the interlayer water (4). The chemical state of charged species in bentonites is further influenced by chemical interaction with the solid phase, such as cation exchange, surface complexation, or solid-solution equilibria. The question of whether such processes may affect not only the retention behavior of positively charged species but also their mass transfer is still a controversial issue discussed in the literature. Evidence pro or contra for such an influence can be found (5-9). As such interactions with cationic species mainly take place in the interlayer water of compacted bentonites, it is still an open question to what extent the diffusion of cations through the interlayer water may contribute to the overall flux through compacted clays. The purpose of the present contribution is to provide new evidence for the importance of interlayer diffusion of cations in compacted swelling clays. As a first step in the direction of investigating the diffusive properties of differently charged cationic species, the diffusion of 22Na and 85Sr has been measured in highly compacted montmorillonite and kaolinite by the classical through-diffusion technique com10.1021/es061908d CCC: $37.00
2007 American Chemical Society Published on Web 12/09/2006
bined with tracer profile analysis (10). The effect of clay thickness and external salt concentration was investigated. 22 Na and 85Sr were chosen as typical representatives of monovalent and bivalent cations sorbing via cation exchange (11-13). The clay minerals were chosen because of the differences in lattice substitutions. Whereas the typical cationexchange capacity of kaolinite is 0.03-0.04 equiv kg-1, in the case of montmorillonite it is of the order of 1 equiv kg-1. There is (almost) no interlayer water in saturated kaolinite. High degrees of compaction were chosen to increase the ratio between interlayer water and free pore water in montmorillonite.
Experimental Section Reagents, Samples, and Analytical Procedures. Reagents of highest purity were obtained from Fluka (Buchs, Switzerland) or VWR (Dietikon, Switzerland). Deionized water was used throughout (Milli-Q water) for preparation of solutions and aqueous extracts. Radioisotopically pure 22Na and 85Sr were obtained from Isotope Products Europe (Blaseg, Germany). Both radionuclides were kept in slightly acid solutions. Radiochemical assays of both isotopes were carried out with a γ-counter (Minaxi-γ, Autogamma 5000 series, Packard). All data were corrected for tracer decay. Montmorillonite from Milos (Greece) was purchased from Su ¨ dchemie. It was equilibrated three times in succession with 1 M NaCl (solid to liquid ratio ∼25 g dm-3) to remove all soluble salts and/or sparingly soluble minerals such as calcite and to convert the clay into the homoionic Na-form. After washing out NaCl by dialysis of the suspensions against deionized water, the clay was freeze-dried. The cationexchange capacity of the final Na-montmorillonite was ∼0.8 equiv kg-1. Kaolinite (KGa-2) was from obtained from the Source Clay Mineral Repository (University of Missouri, Columbia, MO) and was used as received. Diffusion Measurements. Through-diffusion experiments were carried out at 25 °C using confined cylindrical clay plugs (diameter, 2.56 cm; thickness, 5.4 or 10.4 mm) in equipment similar to that as described by Van Loon et al. (14). The wall thickness of the diffusion cells was increased to obtain increased pressure resistance. Variation of sample thickness was achieved by using rings of different lengths. Dry clay samples were compressed directly in the steel rings of the diffusion cells to a target dry bulk density of 1950 kg m-3. After covering each side of the clay sample with a porous stainless steel filter disk (diameter, 2.56 cm; thickness, 1.55 mm; porosity, ∼30%; pore size, 10 µm) and screwing together the diffusion cells, the clays were allowed to saturate for ∼4 weeks with solutions of NaClO4 at the same concentration as used in the subsequent diffusion experiment. Diffusion was started by exchanging the NaClO4 solution in one of the circulation systems by a corresponding solution spiked with the desired radioactive tracer. 85Sr was typically at ∼1000 Bq cm-3 and 22Na at ∼100 Bq cm-3. The spiked solution (the “source reservoir”) was kept unchanged during the whole duration of the diffusion experiment, resulting in a nonstationary upstream boundary condition. The solutions in the other circulation system (the “target reservoir”) were exchanged against fresh NaClO4 solution at regular intervals keeping the tracer concentration at 1400 kg m-3 (18)). It is thus reasonable to define the concentration of B in the interlayer water rather on a mass per volume than on a mass per surface basis. However, the volume of the interlayer water is a free model parameter. For a conceptual approach, it is thus necessary to define it in terms of a measurable quantity. One possibility is to use the concentration of sorbed ions referred to the dry mass of solid (B h, mol kg-1), because the volume of the interlayer space is directly proportional to the mass of solid. Under the assumption that the porosity for interlayer water in compacted clays may be approximated by (cf. eq 2), the relation between cilB and B h is given by
cilB ) B h
Fdb
(5)
with Fdb equal to the dry bulk density of the clay (kg m-3). Equation 4 can thus be rewritten as
Jtot = -D h il
∂B h ∂x
(6)
where
Fdb D h il ) Dil
(7)
D h il is an effective diffusion coefficient formally defined for a dry mass-based concentration gradient and has the units kg m-1 s-1. The tracer gradient used in eq 6, however, cannot be measured in the experiments. As a resort, it may be related to the gradient in the free pore water by a chemical model. As evidenced by the increasing loss of tracer cation from the source reservoir with decreasing external salt concentration, and by the near-linear relationship between modeled Rd values and the external salt concentration (cf. Figure 4A and B), the distribution of the tracer cation between interlayer and external water is assumed to be governed by a cationexchange mechanism. If the clay is conditioned to the homoionic form by the main electrolyte cation (A, with charge zA) present in the external water phase, it may be expected that the diffusive flux of a tracer cation (B, with charge zB) will be dependent on the concentration of the main cation in the external water phase, because there will be competition between A and B for sorption binding sites in the interlayer space, according to the reaction
zBAzA-clay + zABzB a zABzB-clay + zBAzA
(8)
The use of eqs 6 and 8 to describe the transport behavior of B in compacted clays makes much more sense with respect to the experimental observations. D h il is an effective diffusion coefficient depending only on the specific properties of the
equiv kg-1), and brackets denote aqueous molar concentrations. γA and γB are the aqueous phase activity coefficients of A and B. For trace concentrations of cation B, NA approaches a value of 1, and the following relationship between BAKc, Rd (m3 kg-1) and the concentration of a monovalent electrolyte cation (zA ) 1) can be defined: B A Kc
zB(γA)zB zB ) yvRd [A] CECγB
(10)
where yv is a unit conversion factor (yv ) 1000 dm3 m-3). For linear sorption, which is assumed to be valid at low B concentrations, it follows that
∂cB ∂B h dB h ∂cB ‚ ) ) Rd ∂x dcB ∂x ∂x
(11)
Inserting eqs 10 and 11 into eq 6 leads to
CECγB ∂cB Jtot = -D h il ABKc [A]-zB‚ zB ∂x yvzB(γA)
(12)
D h il is independent of [A], while the remaining part of the right-hand side of eq 12 describes the dependence of the concentration gradient of B in the interlayer water as a function of [A] (cf. the analogy with eq 6). Because cD values were evaluated with reference to the concentration gradient in the aqueous phase,
∂cB ∂x
Jtot ) -cD
(13)
the following relationship between cD and D h il can be formally established from a comparison between eqs 12 and 13:
FIGURE 4. Logarithmic representation of the dependence of cD and Rd for 22Na (plot A) and 85Sr (plot B) on the cation concentration in the external water phase [A]. The data are mean values of results from through diffusion experiments and profile analysis; 5- and 10-mm clay thickness. diffusing cation. Based on the presumed analogy between cations present in the interlayer water and cations sorbed via a cation-exchange mechanism, it is reasonable to assume that B h , and thus the concentration gradient of B in the interlayer water, will be a function of the external salt concentration. Whether eq 8 is described using an electrostatic or nonelectrostatic approach is of lesser importance in the present context, because the critical parameters involved, viz. the Donnan exclusion, the distribution of the electrostatic potential, or a selectivity coefficient for cation exchange in the interlayers, are not known for highly compacted clays. As a pragmatic approach, we use a nonelectrostatic description for the distribution of B between the interlayer water and the free pore water. Following the Gaines and Thomas convention (21), a selectivity coefficient, BAKc (dimensions, M(zB-zA)), can be used to quantify the exchange of B for A in eq 8: B AKc
)
(NB)zA[A]zB (γA)zB ‚ (NA)zB[B]zA (γB)zA
(9)
NA and NB (dimensionless) are equivalent fractional occupancies, defined as the equivalents of A (or B) sorbed per unit mass divided by the cation-exchange capacity (CEC, in
c
CECγB D)D h il ABKc [A]-zB yvzB(γA)zB
(14)
For given concentrations of the background electrolyte, yv, zB, D h il, BAKc, and CEC are constant and can be summarized in a lump parameter C. With respect to the experiments presented here, it can be stated that γA and γB cancel from eq 12 in the case of 22Na. For the case of 85Sr, the quotient γB(γA)-zB varies only by ∼7% for ionic strengths between 0.5 and 1.0 M, as calculated using the specific ion interaction theory (22). In view of the relatively large uncertainties of the best-fit parameter values for cD, it is therefore justified to also include γA and γB in C. For the results presented, eq 14 may thus be simplified to
log cD ) log C - zB log[A]
(15)
where -1 -zB C)D h il ABKcCECz-1 B yv γB(γA)
For a given set of fixed parameters and leaving only [A] varying, logcD for 22Na will thus depend linearly on log[A] with a slope of -1, whereas in the case of 85Sr, the slope will be -2. Panels A and B in Figure 4 show that the experimental data support such a postulate. Despite the relatively large overall uncertainties, it is not possible to obtain different integer values for zB from the data other than those proposed. The consistency between postulated and observed values for zB shows that the assumptions made for this evaluation, viz. (i) total tracer transport dominated by the interlayer pathway and (ii) dependence of Rd and consequently of VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
483
interlayer tracer concentration in agreement with eq 10, are justified by the experimental data. Note that the large uncertainties in cD values for 85Sr diffusion are caused by inconsistencies between experiments having a clay sample thickness of 5 or 10 mm. An individual evaluation for each of clay thickness gives a slope of -2 with much smaller standard uncertainties of the data. Another corroboration for this interpretation is given by the experiments with kaolinite. Kaolinite has almost no fixed charges and thus almost no interlayer water. Consequently, it is expected that cD values are fairly independent of the external electrolyte concentration, which has been confirmed by the experiments (cf. Table 3). To the knowledge of the authors, almost no such quantitative relationship between cD for differently charged diffusing metal ions and the external electrolyte concentration has been demonstrated before. The only exception is the work of Muurinen, in which diffusion of Cs+ and Sr2+ was measured in compacted bentonite (23). In contrast to the results presented here, a linear dependency with a slope of -1 of logcD on the logarithmic electrolyte concentration was found for both cations. However, it should be noted that this relationship was established for Sr2+ using only two different electrolyte concentrations, one of which was 0.001 M. Based on the experience gained in the present work, it can be speculated that Sr diffusion through the compacted bentonite at such low external salt concentration is so fast compared to the diffusive resistance of the porous filter plates that the evaluation of reliable diffusion coefficients may be very questionable. Diffusion coefficients may be strongly underestimated when the results are not properly corrected for diffusive resistance of the porous filter plates. The approach chosen in the present work to interpret diffusion data in highly compacted clay systems is related to the concept of surface diffusion or surface migration (e.g., refs 5 and 24-26). However, the present approach differs distinctly with respect to the weighting of Jpw and Jil (cf. eq 3). Whereas it is assumed here that Jpw is negligible as compared to the total flux, Jpw data are conventionally derived from diffusion data measured for HTO using established correlations between diffusion coefficients for HTO and the cation of interest in bulk water. As an example, Jpw for tracer diffusion of Sr through a compacted Illite/smectite sample was calculated, based on the diffusivity of HTO in this clay, as being 44% of the total flux (5). In view of the fact that HTO may also diffuse through the interlayer water (3), it seems doubtful that the diffusion of cations in the actual free pore water can be properly derived from HTO diffusion data. The true flux in the free pore water may be largely overestimated for such an assumption, because these pathways are more restricted with respect to constrictivity and tortuosity than the interlayer water pathways. This is actually reflected by the relatively low effective diffusion coefficients found for anions (27). It is thus obvious that different values for Dil will be calculated using the classical approach (5) and the approach proposed here. Dil values may be calculated from the intercepts (C) of the fit curves in a representation of cD values as a function of the external salt concentration (cf. eq 15) in combination with eq 7. Using BAKc ) 1 and γA ) γB for the exchange of stable Na+ for 22Na+ results in Dil of ∼8 × 10-12 m2 s-1 for the diffusion of 22Na. For 85Sr diffusion, a Dil of ∼2 × 10-12 m2 s-1 is obtained with BAKc ) 2.6 M for the exchange of stable Na for 85Sr. The Rd values calculated from these BAKc (cf. eq 10) deviate from the best-fit values (cf. Table 1) by less than 10% in the case of 22Na. In the case of 85Sr, the best-fit values (cf. Table 2) are higher by a factor of ∼2 than those calculated from BAKc. This also shows that, within the above-mentioned restrictions, the consistency between the concept of single-porous interlayer diffusion and the use of selectivity coefficients for 484
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007
the calculation of cation distribution between in the interlayer and the free pore water is satisfying within the frame of experimental conditions tested. The present work shows that the interpretation of cation diffusion experiments in highly compacted swelling clays in terms of the concentration gradient in the aqueous phase may result in a nonsensical dependence of the effective diffusion coefficients on the salt concentration in the external aqueous phase. An alternative interpretation using an effective diffusion coefficient in the interlayer water (Dil), being independent of the external salt concentration, with a corresponding concentration gradient in the interlayer water is more consistent with the experimental observations. For the experimental conditions tested, this gradient in the interlayer water was assumed to be a function of the external salt concentration. This interpretation explains the observation of the mass flux being a function of the external salt concentration. It is supported (i) by the different slopes observed for 22Na and 85Sr in a logarithmic representation of the conditional diffusion coefficients plotted as a function of the external salt concentration and (ii) by the observation of medium-independent conditional diffusion coefficients in the case of kaolinite, which contains no interlayer water. Thus, the diffusion of cations in highly compacted swelling clays is related to their sorption properties with respect not only to retention but also to the resulting mass flux. The proposed interpretation should in turn not be blindly applied to other experimental conditions. Diffusion of cations via the free pore water may become increasingly important in swelling clays with lower degrees of compaction or in clays in which the interlayer gel pores are not that adjacent as they are in compacted montmorillonite. In such cases, the assumption of Jtot = Jil may no longer hold, and a doubleporous diffusion model would have to be applied in such cases. The present concept may also reach its limits when dealing with cations that rather sorb by surface complexation than by ion exchange. Further work is therefore planned to extend the investigations to such systems.
Acknowledgments We thank the Swiss National Cooperative for the Disposal of Radioactive Waste (Nagra) for partially financing this work. Special thanks goes to F. Berger, J. Kohout, R. Reiser and R. Rosse´ (PSI) for technical support.
Supporting Information Available Evaluation of diffusion parameters from tracer profiles. Breakthrough curves for the diffusion of 22Na in montmorillonite at different concentrations of NaClO4 (Figure S1); tracer profiles of 22Na at different concentrations of NaClO4 (Figure S2); breakthrough curves for the diffusion of 85Sr in montmorillonite and kaolinite (Figures S3 and S4). This material is available free of charge via the Internet at http:// pubs.acs.org.
Literature Cited (1) Pusch, R. The buffer and backfill handbook. Part 1: Definitions, basic relationships, and laboratory methods; SKB Technical Report TR-02-20, SKB: Stockholm, 2002. (2) Bourg, I. C.; Bourg, A. C. M.; Sposito, G. Modeling diffusion and adsorption in compacted bentonite: a critical review. J. Contam. Hydrol. 2003, 61, 293-302. (3) Bourg, I. C.; Sposito, G.; Bourg, A. C. M. Tracer diffusion in compacted, water-saturated bentonite. Clays Clay Miner. 2006, 54, 363-374. (4) Bolt, G. H.; de Haan, F. A. M. Anion exclusion in soil. In Soil chemistry: B. physico-chemical models; Bolt, G. H., Ed.; Elsevier Scientific Publishing Co.: Amsterdam, 1982. (5) Oscarson, D. W. Surface diffusion: Is it an important transport mechanism in compacted clays? Clays Clay Miner. 1994, 42, 534-543.
(6) Conca, J. L.; Apted, M.; Arthur, R. Aqueous diffusion in repository and backfill environment. Mater. Res. Soc. Symp. Proc. 1993, 294, 395-402. (7) Cole, T.; Bidoglio, G.; Soupioni, M.; O’Gorman, M.; Gibson, N. Diffusion mechanisms of multiple strontium species in clay. Geochim. Cosmochim. Acta 2000, 64, 385-396. (8) Molera, M.; Eriksen, T. Diffusion of 22Na+, 85Sr2+, 134Cs+ and 57Co2+ in bentonite clay compacted to different densities: experiments and modeling. Radiochim. Acta 2002, 90, 753760. (9) Jensen, D. J.; Radke, C. J. Caesium and strontium diffusion through sodium montmorillonite at elevated temperature. J. Soil Sci. 1988, 39, 53-64. (10) Shackelford, C. D. Laboratory diffusion testing for waste disposalsa review. J. Contam. Hydrol. 1991, 7, 177-217. (11) Laudelout, H.; Van Bladel, R.; Bolt, G. M.; Page, A. L. Thermodynamics of heterovalent cation exchange reactions in a montmorillonite clay. Trans. Faraday Soc. 1968, 64, 14771488. (12) Van Bladel, R.; Menzel, R. A thermodynamic study of sodiumstrontium exchange on Wyoming bentonite. Proc. Int. Clay Conf. Tokyo 1969, 1, 619-634. (13) Chen, C.; Hayes, K. F. X-ray absorption spectroscopy investigation of aqueous Co(II) and Sr(II) sorption at clay-water interfaces. Geochim. Cosmochim. Acta 1999, 63, 3205-3215. (14) Van Loon, L. R.; Soler, J. M.; Bradbury, M. H. Diffusion of HTO, 36Cl- and 125I- in Opalinus Clay samples from Mont Terri. Effect of confining pressure. J. Contam. Hydrol. 2003, 61, 7383. (15) Williams, A.; De Halleux, B.; Diamondstone, B.; Ellison, S.; Haesselbarth, W.; Kaarls, R.; Mansson, M.; Mo¨ller, H.; Taylor, P.; Thomas, B.; Wegscheider, W.; Wood, R. Quantifying uncertainty in analytical measurement, 1st ed.; EURACHEM Working Group on Uncertainty in Chemical Measurement: London, U.K., 1995. (16) Gold, V. Compendium of chemical terminology: IUPAC recommendations, 2nd ed.; Blackwell Scientific Publications: Oxford, 1997.
(17) Kozaki, T.; Sato, H.; Fujishima, A.; Saito, N.; Sato, S.; Ohashi, H. Effect of dry density on activation energy for diffusion of strontium in compacted sodium montmorillonite. Mater. Res. Soc. Symp. Proc. 1997, 465, 893-900. (18) Kozaki, T.; Fujishima, A.; Sato, S.; Ohashi, H. Self-diffusion of sodium ions in compacted montmorillonite. Nucl. Technol. 1998, 121, 63-69. (19) Li, Y. H.; Gregory, S. Diffusion of ions in sea water and in deepsea sediments. Geochim. Cosmochim. Acta 1974, 38, 703714. (20) Shackelford, C. D.; Daniel, D. E. Diffusion in saturated soil: I: Background. J. Geotech. Eng. ASCE 1991, 117, 467-484. (21) Gaines, G.; Thomas, H. C. Adsorption studies on Clay Minerals II. A formulation of the thermodynamics of exchange adsorption. J. Chem. Phys. 1953, 21, 714-718. (22) Ciavatta, L. The specific interaction theory in evaluating ionic equilibria. Ann. Chim. 1980, 70, 551-567. (23) Muurinen, A.; Rantanen, J.; Penttila¨-Hiltunen, P. Diffusion mechanisms of strontium, cesium and cobalt in compacted sodium bentonite. Mater. Res. Soc. Symp. Proc. 1985, 50, 617624. (24) Jenny, H.; Overstreet, R. Surface migration of ions and contact exchange. J. Phys. Chem. 1939, 43, 1185-1196. (25) de Lopez-Gonzalez, J.; Jenny, H. Diffusion of strontium in ion-exchange membranes. J. Colloid Sci. 1959, 14, 533542. (26) van Schaik, J. C.; Kemper, W. D.; Olsen, S. R. Contribution of adsorbed cations to diffusion in clay-water systems. Soil Sci. Soc. Am. Proc. 1966, 30, 17-22. (27) Muurinen, A.; Penttila¨-Hiltunen, P.; Uusheimo, K. Diffusion of chloride and uranium in compacted sodium bentonite. Mater. Res. Soc. Symp. Proc. 1989, 127, 743-748.
Received for review August 9, 2006. Revised manuscript received October 26, 2006. Accepted October 26, 2006. ES061908D
VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
485
