Hydrothermal decomposition of actinide(IV) oxalates: a new aqueous route towards reactive actinide oxide nanocrystals

Abstract The hydrothermal decomposition of actinide(IV) oxalates (An= Th, U, Pu) at temperatures between 95 and 250 °C is shown to lead to the production of highly crystalline, reactive actinide oxide nanocrystals (NCs). This aqueous process proved to be quantitative, reproducible and fast (depending on temperature). The NCs obtained were characterised by X-ray diffraction and TEM showing their size to be smaller than 15 nm. Attempts to extend this general approach towards transition metal or lanthanide oxalates failed in the 95–250 °C temperature range. The hydrothermal decomposition of actinide oxalates is therefore a clean, flexible and powerful approach towards NCs of AnO2 with possible scale-up potential.


Introduction
Nanocrystals (NCs) represent fundamental building blocks in nanoscience and nanotechnology because of their size and shape dependent properties and have attracted significant interest [1][2][3]. Much has been published on nanomaterials with stable elements, however due to the radioactive nature of the actinides combined with safety issues and restrictions in handling, as well as reduced availability and difficulties in access, in the case of the transuranium elements, less is published on actinide nanoparticles and then practically all on actinide dioxides (AnO 2 ).
An excellent knowledge on the production and properties of nanometric sized AnO 2 is essential for several reasons. They can be used as precursors for the production of materials with special properties, mimic structures which are found in spent nuclear fuels or act as models to study and understand migration in a geological environment. The migration of plutonium in the geosphere proceeds faster than predicted [4,5], indicating a colloidfacilitated transport [6]. As a consequence, a reliable and reproducible method dedicated to the controlled synthesis of actinide-based NCs is highly desirable to perform dedicated studies with respect to mobility of actinides in a well-equipped laboratory.
Different aqueous processes yielding NCs of (over) stoichiometric AnO 2 were also applied [28][29][30]. Among them, oxalate precipitation/decomposition seems to be the most suitable route towards nanograined AnO 2 . However, the so-called "low-temperature oxalate decomposition process" takes place above 600 o C in the case of U IV and Th IV [31]. As a result, the reactivity of such nanopowders would be decreased considerably as, at these temperatures, the aggregation and growth of the nanoparticles starts to proceed.
All the inconveniences presented by the former processes (elaborate methods for organic-driven systems, high temperature use, long reaction time) could be avoided by employing an alternative approach i.e. the hydrothermal decomposition of hydrated actinide oxalates. The synthesis of NCs in hot compressed water under hydrothermal conditions or other solvents is well known and established [32][33][34]. The production of NCs in continuous flow reactors has been applied successfully [35,36]. However, for the synthesis of AnO 2 NCs under hydrothermal conditions and their characterisation nothing has been reported in the literature up to now [12]. Accordingly, in this contribution we report on the synthesis and characterisation of AnO 2 NCs formed under hydrothermal conditions in batch reactors. Attempts to extend this general approach towards transition metal or lanthanide oxalates, however, failed in the 95-250 o C temperature range.

Experimental Procedure
Different amounts (30-100 mg) of An(C 2 O 4 ) 2 ⋅nH 2 O (synthesis procedures described elsewhere) [31,37] were heated at 95-250 °C in 20 ml Teflon autoclaves together with a small volumes (2-5 ml) of water for 5 minutes to several days. After cooling, the autoclaves were opened and the precipitate analysed. The reaction conditions for three representative tests are given in Table 1. In the uranium case, the work under inert atmosphere and addition of stoichiometric amounts of hydrazine were needed to maintain the U IV oxidation state.
XRD analyses were performed on a Rigaku Miniflex 600 diffractometer for the obtained ThO 2 and UO 2 NCs, whereas the PuO 2 NCs were analysed on a Bruker D8 diffractometer equipped with a LinxEye position sensitive detector. Transmission electron microscope analyses (TEM) were performed on a TecnaiG2 (FEI™) 200 kV microscope equipped with a field emission gun, modified during its construction to enable the examination of radioactive samples. The samples for the TEM investigations were prepared by dropping suspended samples on a TEM grid and evaporating the solvent. The elemental analysis of the sample during the TEM study was performed using Electron Energy Loss Spectroscopy (EELS).

Results and Discussions
As can be seen from the XRD data shown in Fig. 1, nanoparticles are formed in all cases; EELS analyses clearly proved the purity of the actinide in the AnO 2 NCs.
The particle sizes for each can be calculated from the XRD data based on the full width at half maximum for six selected peaks in the 2θ range between 25 and 80° for ThO 2 and UO 2 and between 25 and 100° for PuO 2 . They have been determined as 7.2(1.0) nm for ThO 2 , 13.8(1.2) nm for UO 2 , and 4.7(1.0) nm for the PuO 2 nanoparticles, respectively. Concerning the reactivity in the particle formation it turned out that it increases from Th to Pu, i.e. ThO 2 NCs can only be obtained at reaction temperatures of 250 °C, UO 2 NCs are formed at about 110 °C, whereas the PuO 2 NCs are already formed at 95 °C (but the reaction takes several days). Due to the low temperature decomposition, the size of the PuO 2 NCs is smallest, which can be derived obviously from the increased line broadening in Fig. 1. A more detailed size distribution analysis on the dependence on the reaction conditions has still to be concluded, as well as answering the question as to whether the decomposition has to proceed in water or whether other protic solvents or even water free systems are appropriate as well. In the case of the UO 2 NCs it is of real advantage to work under oxygen free conditions as otherwise (especially at low temperatures and long reaction times) oxidation of the U IV to UO 2 (OH) 2 might proceed. Furthermore, it is essential, in  order to control the size of the UO 2 NCs, to work in a nonoxidising environment. From the TEM analyses (Fig. 2) it can be seen that in all of the decomposition reactions NCs are formed. In agreement with the XRD analyses, the particle sizes were determined to be 5.2(1.2) nm and 2.6(0.5) nm for ThO 2 and PuO 2 respectively. In the case of UO 2 , two different distributions could be found. One of them, the most prevalent, had an average particle size of 13.7(3.8) nm, and a second one, less numerous, but representing an important fraction of the sample, for which the average size was found to be 5.5(1.1) nm. From both XRD and TEM analyses it is obvious that the crystallinity of the samples is high, which is shown in the XRD by the diffraction peaks being visible up to high angle and in the high resolution TEM the planes in the crystallites are resolved (Figs. 1 and  2). The shape of the crystals, together with their increased reactivity, enables the consolidation of homogeneous nanostructured mixed oxides (MOX) as intermediates towards very dense nuclear fuels for GenIV reactors.

Conclusions and Perspectives
In conclusion, we present here for the first time the synthesis of highly crystalline, reactive nanograins of AnO 2 (An= Th, U, Pu) through the hydrothermal decomposition of the corresponding actinide(IV) oxalate hydrates. The particles produced in this method are significantly smaller than those produced by conventional thermal decomposition, where temperatures up to 600 o C are needed. Furthermore, the morphology differs as well. The process proved to be quantitative, facile, fast, and reproducible. It contains fewer procedural steps than typical oxalate precipitation/ decomposition processes [31], being suitable for production using a single vessel and under continuous flow. Ongoing experiments focus on the process optimisation, scale-up, and proving its flexibility in the production of MOX nanoparticles. As process temperatures below 250 °C are selective towards the actinide elements it could exhibit a certain application potential for the removal of uranium from contaminated wastewaters at the front-end of the nuclear fuel cycle but more attention has to be addressed to the kinetics and influences of other process parameters like pH or additional salt impurities in the system.