Colloid-borne americium migration in Gorleben groundwater: significance of iron secondary phase transformation.

The mobility of actinides in natural water may be enhanced by colloid-mediated transport. In this context the reversibility of actinide colloid interaction is a key factor. Iron is an element that can generate colloids under conditions found in natural waters. In this paper, the impact of hematite and the low-crystalline precursor 2-line ferrihydrite on colloid-mediated transport of americium(III) is investigated. Am(III)-containing iron colloids are generated from two different approaches, namely contact between the two in aqueous solution or coprecipitation of Am(III) during iron colloid generation. Dissolved organic carbon (DOC), especially humic substances, has a strong influence on the stability of inorganic colloids. In addition, humic substances interfere in the distribution and kinetics of exchange between groundwater and sediments. Four groundwaters from the Gorleben aquifer system are used with DOC concentrations varying between 0.9 and 81.6 mgC/L together with Pleistocene Aeolian quartz sand from this site. Batch and column experiments are conducted under near-natural conditions (Ar + 1% CO2). To study the influence of kinetics, contact times up to one month are studied. The dynamic light-scattering investigations show that the colloidal stability of the 2-line ferrihydrite increases with increasing DOC concentration. The low-crystalline iron colloids have a marginal influence on the Am(III) transport due to reversibility of americium sorption. Contrary to this, the crystalline hematite generated from coprecipitation of Am(III) leads to an increase of unretarded colloid-mediated Am(III) transport up to a factor of almost five. Chemical characterization of these hematite colloids shows that Am(III) is structurally entrapped in the hematite. The distribution of Am(III) and 2-line ferrihydrite between groundwater and sand sediment remained in disequilibrium even after one month. This shows that the kinetics of Am(III) distribution between the different phases (bulk solution/colloidal form/ sediment) is a key issue.

241Am migration in a sandy aquifer studied by long-term column experiments.

The migration behavior of 241Am(III) in a sandy aquifer was studied under near-natural conditions by long-term column experiments of more than 1 year duration. Columns with 50 cm length and 5 cm in diameter were packed with aeolian quartz sand and equilibrated with two different groundwaters having an original dissolved organic carbon concentration (DOC) of 1.1 and 7.2 mg x dm(-3), respectively,from the Gorleben site (Lower Saxony, Germany). In each experiment, 1 cm3 of Am-spiked groundwater ([Am] = 0.2 to 2 micromol x dm(-3)) was injected into the column. The flow rate of the groundwater was adjusted to 0.28 m x d(-1). A small colloid-borne Am fraction was found to elute together with tritiated water. After 414 and 559 days, respectively, the experiments were terminated. Whereas the nonsorbing tracer of tritiated water would have covered a distance of about 350 m in that time period, the maximum of the Am activity was detected between 32 and 40 mm column length. Applying selective dissolution analysis to the sand surface, Am was found to be preferentially bound to iron hydroxide/oxide sites. From this Am distribution, a retardation factor R of about 10(4) was determined and compared to static batch experiments. The Am breakthrough was calculated forthe conditions of the column experiment

A kinetic study of Am(III)/humic colloid interactions.

The interaction kinetics of the Am(III) ion with aquatic humic colloids is investigated under near-natural conditions by column experiments with a sandy aquifer sample rich in humic substancesforthe appraisal of the migration behavior of Am. The association and dissociation kinetics of the Am ion onto and from humic colloids control the migration of colloid-borne Am. As the contact time between Am and humic colloids prior to introduction into a column is increased, the mobility of colloid-borne Am is enhanced and hence the recovery of Am in the effluent increases. On the other hand, an increase of the migration time and residence time in column, respectively, reduces the Am recovery. Considering these experimental results a refined version of the kinetic model KICAM (Kinetically Controlled Availability Model), which suggests different Am binding modes with humic colloids, was developed. Applying KICAM it is possible to predict static and dynamic experiments affected by the kinetically controlled Am/humic colloid interactions over the range of 1 h up to several months. However, to apply these experimental results to long-term conditions, the Am binding scheme as proposed in KICAM needs to be verified. This paper provides, therefore, a basis for a better understanding of the colloid-borne Am migration in porous aquifer systems.