Transcript

Journal of Colloid and Interface Science 324 (2008) 212–215

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Size dispersion and colloid mediated radionuclide transport in a synthetic porousmedia

A. Delos a, C. Walther b,∗, T. Schäfer b, S. Büchner b

a Amphos XXI Consulting S.L. (formerly Enviros Spain S.L.) Pg. de Rub, 29-31, E-08197 Valldoreix, Spainb Forschungszentrum Karlsruhe, Institut für Nukleare Entsorgung, 76021 Karlsruhe, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 28 February 2008Accepted 11 May 2008Available online 16 May 2008

Keywords:ColloidSize dispersionLIBDParticle size distributionTrace analysis

Size dispersion effects during the migration of natural submicron bentonite colloids (<200 nm) througha ceramic column are observed for the first time by laser-induced breakdown detection (LIBD) at ppm(parts per million) mass concentration. Larger size fractions (∼200 nm) arrive prior to smaller size frac-tions (<100 nm) at the column outlet in agreement with model predictions and earlier findings withcarboxylated polystyrene spheres. By addition of trace amounts of americium(III) and plutonium(IV), col-loid mediated transport of these radionuclides is studied. The peak arrival times of Pu-244 and Am-241,as measured by ICP-MS, match the bentonite colloid breakthrough and occur significantly prior to theconservative tracer (HTO) indicating the colloid-borne migration of tri- and tetravalent radionuclides.

© 2008 Elsevier Inc. All rights reserved.

1. Introduction

Solid particles are omnipresent in the aquifer and their size dis-tribution typically extends over many orders of magnitude fromaquatic colloids in the lower nanometer domain to suspended par-ticles of many microns in diameter [1]. Detailed knowledge onthe size distribution is mandatory in many different aspects ofenvironmental and aquatic geochemistry [2,3] which might be il-lustrated by the prominent example of colloid mediated transportof contaminants [4–10]. Metal ions of very low solubility, like thetetravalent actinides, are expected to be removed from solutionby precipitation or surface sorption. However, sorption to colloids,which are mobile in the aquifer, leads to a dramatic enhancementof mobility. Plutonium, which was expected to be immobilizedvery close to its source by precipitation of Pu(IV)hydroxide, wasfound to migrate kilometer distances within years, e.g., close to anuclear test sites in Nevada [6] and in the vicinity of the Mayakreprocessing plant [11].

Hence, colloid mediated transport is considered as one of thekey uncertainties possibly enhancing the mobility of tri- andtetravalent actinides in the far-field of a deep geological repos-itory for high-level nuclear waste, especially in fractured rock(granite) [12]. Although the last decade has brought consider-able progress in the field of colloid detection methods concern-ing colloid size, mass and size distribution determination withinthe nanometer range [13,14], the colloid transport in natural sys-

* Corresponding author.E-mail address: [email protected] (C. Walther).

0021-9797/$ – see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2008.05.017

tems still lacks process understanding [15–17]. The rock matrixin granite systems and the potential process of matrix diffusionas colloid retention mechanism is discussed frequently [12], butdata in this field is scarce. To further understand the migrationof colloid-borne actinides, a field experiment was conducted inthe Grimsel underground laboratory in Switzerland by bringinglaser-induced breakdown detection (LIBD) [18–20] for in situ mon-itoring [21–23]. For this purpose, different dipole geometries andlengths in a shearzone (MI shearzone) were tested to elucidatethe mobility of colloids and associated metals under the chosenhydraulic conditions [21]. Bentonite colloids (size distribution be-tween ca. 100–180 nm) were injected into the fracture. The colloidcontent and mean colloid size of the outflow was measured byLIBD, time resolved, and the metal concentration was followed byoff-line ICP-MS analysis. The peak maxima of the colloid and metalbreakthrough curves (Th(IV), Hf(IV), and Tb(III)) appeared about10 min ahead of the nonsorbing (conservative) tracer uranine. Theshift suggested the existence of size and/or charge exclusion effectstypical to colloid migration. However, since only the mean colloiddiameters were measured, the size exclusion effect could not bequantified directly.

Recent advances of the LIBD technique [14,24], allow us to mea-sure the particle size distribution (PSD) rather than just a meansize of the colloids and, as a consequence, the size dispersion candirectly be observed. In order to avoid uncertainties arising fromill defined conditions in natural matrices, for a first experiment wechose synthetic porous material of a well-defined pore character-istics. The migration of previously characterized natural polydis-perse colloids, derived from a potential bentonite backfill material(FEBEX), was then observed by LIBD.

A. Delos et al. / Journal of Colloid and Interface Science 324 (2008) 212–215 213

Fig. 1. Bentonite colloid size distribution of injection cocktail used throughout this study. Sample of injection cocktail was diluted 1:100. Colloid number distribution (left)and calculated colloid surface per size class assuming spherical geometry of colloids (right).

2. Materials and methods

2.1. LIBD

Only a short summary of the LIBD technique is given here. For adetailed description on laser-induced breakdown refer to [25] andfor a description on LIBD to a recent review [26]. The methodis based on plasma formation by a focused (d ∼ 7 μm) pulsedlaser beam due to dielectric breakdown [19,25,27,28], a processwhich works much more efficient in solids than in liquids. Hence,a plasma is not ignited in pure water but only when a colloidenters the focal region and thus single colloids are counted. Therelative number of events per number of laser pulses provides ameasure of colloid number density (e.g., 100 plasmas for 1000laser pulses results in a breakdown probability, BDP, of 10%). Inthe present work, the plasma is observed by the detection of theacoustic shock wave generated by its rapid expansion [20,29,30].The BDP is measured for increasing pulse energy (i.e., increasing ir-radiance) which results in so called s-curves (Fig. 2). The thresholdof these curves is a measure of particle size and the slope of thecurve scales with particle concentration. Using reference colloidsof well-defined diameter, a calibration curve is obtained, whichrelates breakdown thresholds to particle size [18]. A recently de-veloped data evaluation scheme for s-curve analysis [24] allows usto measure the particle size distribution of colloids between 15and 400 nm at concentrations down to about 104 particles/mL.

2.2. Column migration experiments

For the colloid migration experiments, a Plexiglas column withthe dimensions of 80 mm length and 40 mm diameter was filledwith ceramic material (Soilmoisture Type B02M2) in the size frac-tion 2–4 mm. The porous ceramic material was characterized bymercury intrusion porosimetry and showed a matrix pore size dis-tribution in the range from 30 to 700 nm with a maximum around250 nm. The migration experiments were set-up in an argon glovebox. The injection cocktail contained 20 mg/L Febex bentonite col-loids suspended in Grimsel ground water [31] with the actinidesAm-241 (4.9 × 10−10 mol/L) and Pu-244 (7.4 × 10−10 mol/L) as

well as tritiated water (HTO) used as conservative tracer. Bothradionuclides, Am(III) and Pu(IV), are in the injection cocktail ben-tonite colloid associated with 84 ± 8% (Pu-244) and 99 ± 8% (Am-241), respectively. The flow rate in the migration experiments wasset to 2 mL/min and a total cocktail volume of 1 mL was injectedvia an injection device into the column inlet [32]. Fractions col-lected at the column outlet by a fraction collector (Gilson Ltd. TypeF204) were analyzed off-line by liquid scintillation counting (LSC),ICP-MS and LIBD s-curve analysis.

3. Results and discussion

The colloid size distribution of the injection cocktail determinedby LIBD s-curve analysis is shown in Fig. 1. The number weightedsize distribution of the Febex bentonite colloidal fraction showstwo maxima around 10–30 and 100–200 nm, respectively. Tak-ing into account the density of clays (ρ = 2.7 g/cm3) as well asassuming spherical geometry, the surface area and mass per sizeclass can be estimated. Based on these assumptions, the total massis 18.7 μg/mL compared to 20 μg/mL determined by gravity and Alconcentration analysis (ICP-MS) using the structural formula givenin [31]. The calculated total surface area is in the range of 10–11 m2/g.

A compilation of the s-curves measured for fractions sampled inthe colloid breakthrough of the migration experiment are shown inFig. 2. At small elution volumes (V elu = 12 mL) or at short times,respectively, only a very low number of colloids is detected (Fig. 2,background). Close to the maximum of the breakthrough curve(V elu = 40 mL), the s-curve is shifted noticeably to the left, indi-cating the presence of large colloids. The corresponding particlesize distribution (Fig. 2, bottom left), which is obtained from the s-curve analysis, is centered around 100 nm. With increasing elutionvolume the thresholds of the s-curves shift to the right indicatinga decreasing mean particle size. Close to the end of the colloid-breakthrough peak at V elu = 79 mL more and more small colloidsare eluted (Fig. 2, bottom right).

HTO and colloids are recovered quantitatively within the analyt-ical errors with 104 ± 5 and 97 ± 5%, respectively. The comparisonof the HTO and colloid mass breakthrough curves (Fig. 3) shows a

214 A. Delos et al. / Journal of Colloid and Interface Science 324 (2008) 212–215

Fig. 2. Raw data s-curves of fractions taken in the colloid breakthrough (top) and two examples of particle size distributions obtained from the s-curves (bottom). Largecolloids are eluted first (left) smaller ones later (right).

peak arrival time of colloids significant earlier than that of the con-servative tracer HTO. This behavior has been observed in the liter-ature [32,33] and is interpreted as a size chromatography effect.The peak arrival times of Pu-244 and Am-241 correspond to thecolloid breakthrough showing a colloid mediated transport. How-ever, the recoveries for both radionuclides are significantly lowerwith 46–53% compared to the colloid recovery. These results showclearly the reversibility of Am and Pu binding to montmorillonitecolloids under the given experimental conditions. Pu(IV) sorptionreversibility, as found in this study, was also observed in clay-richformations favoured as deep-geological host rocks for the storageof high-level nuclear waste [34] and earlier laboratory studies onsorption reversibility by Geckeis and coworkers [21].

For 14 samples between (V elu = 4–350 mL) the particle sizedistributions were measured by LIBD and are visualized in a con-tour plot (Fig. 4). Darker shades indicate higher particle number.This analysis reveals an earlier arrival of larger bentonite size frac-tions (100–200 nm) at the column outlet, whereas smaller sizefractions (50–70 nm) are retarded. These findings cooperate ear-lier modelling results [35] and laboratory investigations on mix-tures of monodisperse synthetic colloids (i.e., carboxylated mi-crospheres) [33]. Detailed studies using sedimentation field-flowfractionation coupled with ICP-MS (Sd FFF-ICP–MS) have already

demonstrated the depth-dependent trends of mobile colloid sizesin soil profiles ranging from 100–500 nm [36]. However, themethod mentioned above as well as split-flow thin-cell (SPLITT)fractionation is limited to colloid sizes >100 nm [37,38]. Only flowFFF (FlFFF) has been proven to give colloid size distribution in-formation in the sub-100 nm range including its coupling withsensitive detection techniques as ICP-MS [33,39,40]. To our knowl-edge, the results presented here demonstrate for the first time thattransport induced chromatographic size separation effects can beresolved even for polydisperse natural colloids in the sub-100 nmrange using LIBD s-curve analysis. The results presented hereshowing the applicability of the LIBD s-curve method to naturalpolydisperse samples without pre-treatment steps opens the doorto investigate colloidal transport in complex shearzone geometrieshaving double peak conservative tracer breakthrough curves, there-fore indicating multiple migration pathways. Planned experimentsin the framework of the Grimsel Test Site (GTS) Phase VI the in-ternational Colloid Formation and Migration (CFM) project withpartners from Japan (JAEA, AIST, and CRIEPI), Switzerland (NAGRA),Sweden (SKB) and Germany (FZK-INE) will focus on the availabilityof such features for colloid migration and the potential size chro-matography along these flowpaths giving new insides into colloidretention processes.

A. Delos et al. / Journal of Colloid and Interface Science 324 (2008) 212–215 215

Fig. 3. Comparison of the conservative (nonreactive) tracer HTO (broken line) andthe Febex bentonite colloids breakthrough curve ("). Furthermore, the break-through curves of Am-241 (✩) and Pu-244 (E) are plotted showing a colloidcomparable arrival time of the peak maximum.

Fig. 4. Contour map of colloid size distribution in the breakthrough curve derivedfrom LIBD s-curve fitting. The plotted contour lines give the colloid concentrationin ppt. The dashed line indicates that larger size fractions arrive prior to smallersize fraction at the column outlet. Based on the peak maximum for the conserva-tive tracer HTO measured at 59.4 mL the breakthrough of the larger size fractioncorresponds to a retardation factor Rf of 0.60 and the smaller size fraction to a Rfof 0.72, respectively.

Acknowledgments

The authors thank H. Geckeis for valuable discussions andF. Geyer and C. Walschburger for the ICP-MS analysis. The resultspresented were partly supported by Network of Excellence (NoE)ACTINET (project 01-01) and the Integrated Project (IP) FUNMIGunder the contract number FP6-516514. The results presented in

the present article were collected during the Ph.D. thesis of AnneDelos granted by Andra (French National Agency for Nuclear WasteDisposal).

References

[1] O. Atteia, D. Perret, T. Adatte, R. Kozel, P. Rossi, Environ. Geol. 34 (1998) 257–269.

[2] P.A. Kralchevsky, K.D. Danov, N.D. Denkov, Chemical physics of colloid systemsand interfaces, in: K.S. Birdi (Ed.), Handbook of Surface and Colloid Chemistry,CRC Press, New York, 1997, pp. 333–477.

[3] T. Kim, C. Lin, Y. Yoon, J. Phys. Chem. B 102 (1998) 4284–4287.[4] C. Degueldre, Mater. Res. Symp. Proc. 465 (1997) 835–846.[5] M. Flury, J.B. Mathison, J.B. Harsh, Environ. Sci. Technol. 36 (2002) 5335–5341.[6] A.B. Kersting, D.W. Efurd, D.L. Finnegan, D.J. Rokop, D.K. Smith, J.L. Thompson,

Nature 397 (1999) 56–59.[7] J.I. Kim, Mater. Res. Soc. Bull. 19 (1994) 47–53.[8] T. Klein, R. Niessner, Vom Wasser 87 (1996) 373–385.[9] J. McCarthy, C. Degueldre, Sampling and characterization of colloids in

groundwater for studying their role in contaminant transport, in: J. Buffle,H.P.v. Leeuwen (Eds.), Environmental Particles, Lewis, Chelsea, MI, 1993.

[10] J.N. Ryan, M. Elimelech, Colloids Surf. 107 (1996) 1–56.[11] A.P. Novikov, S.N. Kalmykov, S. Utsunomiya, R.C. Ewing, F. Horreard, A. Merku-

lov, S.B. Clark, V.V. Tkachev, B.F. Myasoedov, Science 314 (2006) 638–641.[12] SKB, RETROCK Project (Treatment of geosphere retention phenomena in safety

assessment models), SKB report R-04-48, Sweden, 2005.[13] C. Walther, Colloids Surf. A Physicochem. Eng. Aspects 217 (2003) 81–92.[14] C. Walther, H.R. Cho, T. Fanghänel, Appl. Phys. Lett. 85 (2004) 6329–6331.[15] R. Kretzschmar, T. Schäfer, Elements 1 (2005) 205–210.[16] J.Y. Chen, C.H. Ko, S. Bhattacharjee, M. Elimelech, Colloids Surf. A Physicochem.

Eng. Aspects 191 (2001) 3–15.[17] S. Bhattacharjee, C.H. Ko, M. Elimelech, Langmuir 14 (1998) 3365–3375.[18] C. Walther, C. Bitea, W. Hauser, J.I. Kim, F.J. Scherbaum, Nucl. Instrum. Methods

B 195 (2002) 374–388.[19] F.J. Scherbaum, R. Knopp, J.I. Kim, Appl. Phys. B 63 (1996) 299–306.[20] T. Kitamori, K. Yokose, K. Suzuki, T. Sawada, Y. Goshi, Jpn. J. Appl. Phys. 27

(1988) L983–L985.[21] H. Geckeis, T. Schäfer, W. Hauser, T. Rabung, T. Missana, C. Degueldre, A. Mori,

J. Eikenberg, T. Fierz, W.R. Alexander, Radiochim. Acta 92 (2004) 765–774.[22] W. Hauser, H. Geckeis, J.I. Kim, T. Fierz, Colloids Surf. A 203 (2002) 37–45.[23] A. Möri, W.R. Alexander, H. Geckeis, W. Hauser, T. Schäfer, J. Eikenberg, T. Fierz,

C. Degueldre, T. Missana, Colloids Surf. A 217 (2003) 33–47.[24] C. Walther, S. Büchner, M. Filella, V. Chanudet, J. Colloid Interface Sci. 301

(2006) 532–537.[25] L.J. Radziemski, D.A. Cremers, in: B.J. Thompson (Ed.), Laser Induced Plasmas

and Applications. Optical Engineering, Dekker, Rochester, 1989.[26] J.I. Kim, C. Walther, Laser-induced breakdown detection, in: J. Lead, K. Wilkin-

son (Eds.), Environmental Colloids and Particles: Behaviour, Separation andCharacterisation, Wiley, West Sussex, 2007, pp. 555–612.

[27] C.A. Sacchi, J. Opt. Soc. Am. B 8 (1991) 337–345.[28] F. Docchio, A. Avigo, R. Palumbo, Europhys. Lett. 15 (1991) 69–73.[29] W. Hauser, R. Götz, Druckwellensensor, Patent 19602048C2, Germany, DE, 1999.[30] M. Saito, S. Izumida, K. Onishi, J. Akazawa, J. Appl. Phys. 85 (1999) 6353–6357.[31] M.V. Villar, P.L. Martin, M. Pelayo, B. Ruiz, P. Rivas, E. Alonso, A. Lloret, X. Pin-

tado, A. Gens, J. Linares, F. Huertas, E. Caballero, C. Jimenz de Cisneros, J. Obis,A. Perez, J. Velasco, FEBEX Bentonite: Origin, Properties and Fabrication ofBlocks (Full-Scale Engineered Barriers Experiment in Crystalline Host Rock), En-resa, Madrid, 1998.

[32] R. Artinger, B. Kienzler, W. Schüssler, J.I. Kim, J. Contam. Hydrol. 35 (1998)261–275.

[33] T. Schäfer, H. Geckeis, M. Bouby, T. Fanghänel, U. Th, Radiochim. Acta 92 (2004)731–737.

[34] C. Latrille, J. Ly, M. Herbette, Radiochim. Acta 94 (2006) 421–427.[35] J.N. Ryan, R.A. Ard, R.D. Magelky, M. Elimelech, N. Sun, N.Z. Sun, Colloid mo-

bilization and transport in contaminant plumes: Field experiments, laboratorystudies, and modeling, USEPA—United States Environmental Protection Agency,1999.

[36] J.F. Ranville, D.J. Chittleborough, R. Beckett, Soil Sci. Soc. Am. J. 69 (2005) 1173–1184.

[37] A. De Momi, J.R. Lead, Environ. Sci. Technol. 40 (2006) 6738–6743.[38] J.R. Lead, K.J. Wilkinson, Environ. Sci. Chem. 3 (2006) 159–171.[39] M. Bouby, H. Geckeis, T.N. Manh, J.I. Yun, K. Dardenne, T. Schafer, C. Walther,

J.I. Kim, J. Chromatogr. A 1040 (2004) 97–104.[40] H. Geckeis, T.N. Manh, M. Bouby, J.I. Kim, Colloids Surf. A Physicochem. Eng.

Aspects 217 (2003) 101–108.


Top Related