PII S0016-7037(99)00246-X Experimental and modelling studies of caesium sorption on illite CHRISTOPHE POINSSOT, 1,2, *BART BAEYENS, 1 and MICHAEL H. BRADBURY 1 1 Paul Scherrer Institut, Laboratory for Waste Management, CH-5232 Villigen PSI, Switzerland 2 Commissariat a ` l’Energie Atomique, CEA-Saclay, DCC/DESD/SESD, Laboratory for Migration and Solid Geochemistry, F-91191 Gif-sur-Yvette Cedex, France (Received October 15, 1998; accepted in revised form April 29, 1999) Abstract—A natural illite (illite du Puy) was purified and converted to the homo-ionic Na form. The conditioned Na–illite was characterised in terms of its mineralogy, chemical inventory, and surface properties. The structural formula was determined from EDS analyses (SEM/TEM) and bulk chemistry. A cation exchange capacity of 127 mEq/kg was determined by the Na isotope dilution method at neutral pH. The sorption of Cs was measured as a function of NaClO 4 background electrolyte concentration (1.0, 0.1 and 0.01 M), Cs concentration and pH in the range 3 to 10. Before obtaining these measurements the kinetics of Cs uptake were determined at initial concentrations of 2  10 -8 M and 7  10 -5 M, representing the extremes of the range investigated, and was found to be concentration dependent. The supernatant solutions after centrifugation were analysed for major cations in all of the sorption tests. A two-site cation exchange model was developed to describe the sorption of Cs over the whole range of experimental conditions. The two-site types were termed frayed edge sites, FES (high affinity/low capacity) and type II sites (low affinity/high capacity). At low NaClO 4 concentrations, Cs sorption decreased at pH values less than neutral. This was interpreted in terms of competitive effects from H, and K released by the partial dissolution of illite, which cannot be avoided at low and high pH values. Selectivity coefficient values for Cs–Na, Cs–K, K–Na, and H–Na exchange equilibria on the FES sites, and Cs–Na exchange on the type II sites are given for illite together with the corresponding site capacities. Copyright © 1999 Elsevier Science Ltd 1. INTRODUCTION Since the 1950s, the nuclear industry has accumulated substan- tial amounts of radioactive waste. In most cases the intention is to dispose of such waste in underground repositories. The requirement is to isolate the radioactive elements from the biosphere for very long periods of time, up to a million years and longer in the case of high level waste. During such long times, the confinement of the radioelements will be ensured by the presence of several technologic and natural barriers be- tween the radioelements and the biosphere, technological near- field barriers and the geologic medium. At some point in time in the far future, certain radionuclides may have penetrated the near-field confinement and begun to migrate in the groundwater of the host rock where their movement should be (strongly) retarded by a number of physical and chemical processes. One of the most important of these is sorption, particularly on clay mineral surfaces. The uptake of radionuclides by immobile solids is one of the main pillars upon which performance assessments for under- ground repositories are based. Clearly, an understanding of the retention processes occurring at the mineral–solution interface over a representative range of relevant conditions (pH, aqueous composition, mineralogical composition of the host rock, tem- perature) is of prime importance for making a credible safety case. One approach to meeting such a demanding goal is to first understand sorption processes on key mineral components of the natural rock. In many rock types under consideration as potential host formations for repositories, clay minerals are particularly important sinks for radionuclides because of their abundance in natural systems, and their favourable high sorp- tion properties. Cs is an important radionuclide for several reasons: it exhib- its almost unlimited solubility, its inventory in radioactive waste is significant, 135 Cs is extremely long-lived, and in addition 137 Cs has been introduced in the environment through weapons testing and by nuclear accidents (Tchernobyl). Fur- thermore, because of its chemical similarity to K, Cs is readily assimilated by terrestrial and aquatic organisms. The bioavail- ability of Cs in natural systems depends to a large extent on the sorption properties of the solid phases (see for example, Ker- pen, 1986). The sorption of Cs has been extensively studied on a wide variety of sorbents reflecting its importance as a radio- nuclide relevant for safety assessments (Komarneni, 1979; Gill- ham et al., 1980; Komarneni and Roy, 1980; Shiao et al., 1981; Baeyens, 1982; Torstenfelt et al., 1982; Brouwer et al., 1983; Gru ¨tter et al., 1986; 1990; 1994; Cremers et al., 1988; Lieser and Steinkopff, 1989; De Preter, 1990; De Preter et al., 1991; Kaukonen et al., 1993; Wauters et al., 1996a; b; c; Staunton and Roubaud, 1997). In particular, it has been demonstrated that at low concentrations Cs sorbs strongly on micaceous minerals such as illite (Sawhney, 1972; Francis and Brinkley, 1976). In this article, the sorption of Cs as a function of equilibrium Cs concentration, ionic strength, and pH on a conditioned illite, the “illite du Puy” from the Massif Central region of France (Gabis, 1958), is described. Also the kinetics of Cs uptake on the two main sorption sites have been investigated as a function of pH. A thorough physicochemical characterisation of the illite and a careful study of illite reactivity during the sorption *Author to whom correspondence should be addressed (poinssot@ azurite.cea.fr). Pergamon Geochimica et Cosmochimica Acta, Vol. 63, No. 19/20, pp. 3217–3227, 1999 Copyright © 1999 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/99 $20.00 + .00 3217