Impact of the uranium (VI) speciation in mineralised urines on its extraction by calix[6]arene bearing hydroxamic groups used in chromatography columns

Abstract

Actinides determination in urine samples is part of the analyses performed to monitor internal contamination in case of an accident or a terrorist attack involving nuclear matter. Mineralisation is the first step of any of these analyses. It aims at reducing the sample volume and at destroying all organic compounds present. The mineralisation protocol is usually based on a wet ashing step, followed by actinides co-precipitation and a furnace ashing step, before redissolution and the quantification of the actinides by the appropriate techniques. Amongst the existing methods to perform the actinides co-precipitation, alkali-earth (typically calcium) precipitation is widely used. In the present work, the extraction of uranium(VI), plutonium(IV) and americium(III) from the redissolution solutions (called “mineralised urines”) on calix[6]arene columns bearing hydroxamic groups was investigated as such an extraction is a necessary step before their determination by ICP–MS or alpha spectrometry. Difficulties were encountered in the transfer of uranium(VI) from raw to mineralised urines, with yield of transfer ranging between 0% and 85%, compared to about 90% for Pu and Am, depending on the starting raw urines. To understand the origin of such a difficulty, the speciation of uranium (VI) in mineralised urines was investigated by computer simulation using the MEDUSA software and the associated HYDRA database, compiled with recently published data. These calculations showed that the presence of phosphates in the “mineralised urines” leads to the formation of strong uranyl-phosphate complexes (such as UO2HPO4) which compete with the uranium (VI) extraction by the calix[6]arene bearing hydroxamic groups. The extraction constant of uranium (VI) by calix[6]arene bearing hydroxamic groups was determined in a 0.04molL−1 sodium nitrate solution (logK=4.86±0.03) and implemented in an extraction model taking into account the speciation in the aqueous phase. This model allowed to simulate satisfactorily the experimental uranium extraction data and to support the preliminary conclusions about the role of the phosphates present in mineralised urines. These calculations also showed that the phosphate/calcium ratio is a key parameter as far as the efficiency of the uranium (VI) extraction by the calix[6]arene columns is concerned. It predicted that the addition of CaCl2 in mineralised urines would release uranium (VI) from phosphates by forming calcium (II)-phosphate complexes and thus facilitate the uranium (VI) extraction on calix[6]arene columns. These predictions were confirmed experimentally as the addition of 0.1molL−1 CaCl2 to a mineralised urine containing naturally a high concentration of phosphate (typically 0.04molL−1) significantly increased the percentage of uranium (VI) extraction on the calix[6]arene columns.