Production of powder targets for neutron-induced cross section measurements

. Nuclear powder targets for neutron-induced cross section measurements were prepared by pressing. The choice of the production technique was related to the type of nuclear experiment and the quality of the powder. This paper describes the production process of unsintered compacts of 94 Mo, 95 Mo, 96 Mo metal powders and 239 PuO 2 powder with a thickness as low as possible by uniaxial pressing of powders. Special attention goes to the difficulties encountered during the preparation process of 239 PuO 2 pellets. Investigation on the presence of impurities and so the quality of the PuO 2 material was performed by scanning electron microscopy with energy dispersive X-ray spectroscopy, by inductively coupled plasma mass spectrometry, X-ray diffraction and by thermal analysis.


Introduction
Targets for neutron-induced cross-section measurements are prepared according to the type of nuclear experiment, the requested target characteristics, final quality of the sample, material and techniques available [1]. In case thick targets are required with an areal density up to g·cm -2 , preferably a metallic disc is used. In case the isotope of interest is only available as a powder and the conversion into a metal is not possible, the powder is compacted into mechanically stable pellets [2].
Compaction is a widely applied method. Powder compacts can be obtained by uniaxial pressing, isostatic pressing, metal injection moulding or by less used processes such as powder rolling, extrusion and dynamic or explosive compaction. Often a binder is added to favour densification during the compaction process and a lubrication to decrease friction with the inner wall of the press tool. The compacts are subsequently sintered to induce consolidation and modify the porosity as well as the microstructure irreversibly from contacting particles resulting in compacts with a density higher than 95% of the theoretical density (TD) [3][4][5][6][7][8].
This paper describes the preparation at JRC Geel of unsintered pellets of 94 Mo, 95 Mo, 96 Mo metal powders and 239 PuO2 powder with a mass of 2 g and with a thickness as low as possible with a maximum of about 1 mm by uniaxial pressing of powders. To avoid addition of impurities and enable simple recovery of limited and valuable material, pellets were produced without adding fluids or binders or lubricants. * Corresponding author: Goedele.SIBBENS@ec.europa.eu The Mo and PuO2 pellets will be used as targets in neutron-induced cross section measurements. The powder compacts of Mo isotopes are equally important in nuclear astrophysics, nuclear energy application, particularly for GenIV reactors, and in fundamental physics [9,10]. The production of respective targets is supporting one of the main requests for the H2020 EURATOM Project Supplying Accurate Nuclear Data for energy and non-energy Applications (SANDA) [11]. Neutron-induced cross sections of 239 Pu are of great value as 239 Pu is one of the "big three" nuclei ( 235 U, 238 U, 239 Pu) highlighted by the Collaborative International Evaluated Library Organisation (CIELO) Project [12].
The production of the 239 PuO2 pellets for the measurement of 239 Pu (n,n'γ) and (n,2nγ) cross sections within the JRC-wide open access programme EUFRAT [13,14] got special attention because of mechanical instability and non-homogeneity of the pellets and corrosion of the press tools. For better understanding of these problems, the Pu material was characterized at JRC Karlsruhe for impurities by scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDX), by inductively coupled plasma mass spectrometry (ICP-MS), X-ray diffraction (XRD) and by thermal analysis (TA).

Production process
The aim of compaction was to obtain a pellet with sufficient strength to withstand further handling operations without loss of material. The press tool was made of hardened steel with the desired diameter and the corresponding lower and upper dies machined to close tolerances. The maximum filling height and the maximum allowed pressing force, 1.0 kN·mm -2 for a diameter above 9.0 mm, was respected in function of the diameter of the die. The press tool was used in combination with the Specac (15-t pressure load) or Retsch (25-t pressure load) press. In case of Pu, the P/O/Weber manual 2-Column Laboratory Press, Model PW 30-BOX, enclosed in a glove box, was used. All consisted of a hydraulic lab press, an external hand pump and had the same standard operation procedures.
The pelletizing process contained three main stages: (1) die filling, (2) compaction and (3) ejection. A quantitative amount of the powder was inserted directly into the matrix. The required pressure force was generated by hand pumping the hydraulic fluid (oil) to raise the piston and reaching the designated press force.
Prior to the production of the 98.97 at% 94 Mo, 95.40 at% 95 Mo, and 95.90 at% 96 Mo metal powder pellets, tests were done with natural Mo metal powder of grain size < 5 µm and of 350 µm, and prior to the 239 PuO2 pellets CeO2 and natural UO2 were used. The tests were performed to find the optimum press parameters for a mechanically stable pellet with a thickness as low as possible and a maximum of about 1 mm. Finally, the Mo pellets were pressed with 2 g of each isotope and a die diameter of 20 mm because the tests with a diameter of 30 mm resulted in broken pellets.
The 239 PuO2 pellets were produced from different ORNL (Oak Ridge National Laboratory) batches: batch n° 716 (99.97% wt% 239 Pu), batch n° 1756 (99.90% wt% 239 Pu) and batch n° 1756p (batch n° 1756 after purification to remove Am). Batch n° 1756 originates from the same mother batch as batch n° 716 is, but less enriched in 239 Pu. The purification of batch n o 1756 at SCK CEN in Belgium was done by peroxide precipitation, re-dissolution of the plutonium peroxide in nitric acid, oxalate precipitation and calcination for 2h at 735 °C under air. This temperature is above the expected completion of plutonium oxalate decomposition and below the pre-sintering stage. Temperature, atmosphere and hold time were chosen to ensure complete decomposition of the oxalate, yet good compression characteristics of the PuO2 powder. Sinterability was not a goal, but the conditions were expected to enable good densification [15]. The test with 2 g of natural UO2 and a diameter of 30 mm showed that the pellet would be too fragile for handling. Therefore, instead of one pellet with a diameter of 30 mm, four 239 PuO2 pellets with a diameter of 12 mm were produced.

Characterization
All pellets were characterized for mass by weighing and for thickness with a digital thickness gauge. The density, calculated from the mass, the thickness and the inner diameter of the die, was compared with the TD.
The bad strength of the Pu pellets produced from the first batch n° 716, the heterogeneity of the colour and the corrosion of the press tools raised questions related to the presence of impurities. Therefore, a PuO2 pellet (n° 050) and part of the original powder of batch n° 716 (n° 051) were analysed by ICP-MS, SEM-EDX, XRD, and differential thermal analysis and thermogravimetric analysis (DTA-GTA).
ICP-MS was done with the 'nuclearised' ElementXR instrument (Thermo Fisher, Bremen, Germany). About 140 mg of the pellet and about 370 mg of the powder were dissolved, each in 30 mL concentrated HNO3 + 0.1M HF on a hotplate at 90 °C for 48 hr. An important remark is that the samples did not completely dissolve. The supernatant was diluted by a factor of about 10000. The complete mass spectrum was measured from 6 till 248 amu (atomic mass unit).
The elemental composition of the pellet, some particles and the powder were investigated by SEM-EDX at different positions of the material. Analyses were obtained in a Philips/FEI™ XL40 SEM operated at 25 kV, equipped with a SAMx EDX.
The crystal structure was studied by XRD to check the purity of the Pu sample and identify the crystalline phases and composition information. XRD analysis was performed on a Bruker D8 Bragg-Brentano advanced diffractometer (Cu Kα1 radiation) equipped with a Lynxeye linear position sensitive detector and installed inside a glove box under inert atmosphere. The powder diffraction patterns were recorded at room temperature using a step size of 0.01316° with an exposure of 4 s across the angular range 10°⩽ 2θ ⩽ 120°. The operating conditions were 40 kV and 40 mA. Phase analyses were done using the powder diffraction (PDF2) and the inorganic crystal structure database (ICSD). Quantification of the sample was performed by Rietveld refinement using Highscore Plus software.
Thermogravimetry differential TA (TG-DTA) was carried out under flow of oxygen (99.99%) and argon (99.999%) at a heating rate of 10 °C·min -1 on a temperature range of Troom till 1400 °C. The apparatus used was NETZSCH Simultaneous Analyzer STA 449 Jupiter. The weight loss of the material was investigated along heating by DTA-GTA to link it to the potential volatile impurities present in the sample.

Results and discussion
To produce a pellet with a thickness as low as possible and not exceeding a maximum of 1 mm, the diameter was chosen in function of the amount of material required for the experiments and the compact strength.
Pressing 2 g natural Mo and natural UO2 with a diameter of 30 mm resulted in pellets with a thickness from 0.5 mm to 0.7 mm. The pellets did not have good strength and broke during the ejection phase. This was not the case for a diameter of 20 mm. Mechanically stable 94 Mo, 95 Mo, 96 Mo pellets could be prepared with a thickness from 1 mm to 1.1 mm. At the same pressure force of 160 kN, the natural Mo pellets had a thickness of 0.9 mm. The grain size did not influence the pressing process. Fig. 1 shows a picture of a 95 Mo pellet with a mass of 1.98 g, a diameter of 20 mm and a thickness of 1.1 mm after ejection on top of the press die. The pellet has a good strength, has sharp borders and glossy surfaces and can be further handled without loss of Mo.
The tests applying the envisaged press parameters were successful with CeO2 and natural UO2, resulting in stable pellets of desired diameter of 12 mm and thickness of maximum 1 mm. However, when applying the same press parameters using different 239 PuO2 materials, this behaviour was not fully reproducible.
The first PuO2 pellet of batch n° 716 with a diameter of 12 mm could be produced with a good strength while the second, third and fourth pellets lost particulate material. All four had a thickness of about 0.75 mm and showed coloured spots. The press tools that were touching the PuO2 were corroded after each pressing process. This could be due to impurities in the powder. The press procedure applied to PuO2 of batch n o 1756 was successful. As shown in Fig. 2a, this material could be pressed into a pellet with a diameter of 12 mm and good strength. Because of the presence of 241 Am, the four pellets could not be used for the envisaged experiment. The pellets produced with the PuO2 purified for Am, batch n o 1756p, showed the same behaviour as the ones from batch n° 716, namely the first pellet had a good strength (Fig.2b) and thereafter the pellets lost particulate material (Fig.2c). In addition, the press tools that were touching the PuO2 were corroded after each pressing process. The observed variability in the colour of the PuO2 is well known, and while PuO2 is normally olive green, samples of yellow, buff, khaki, tan, slate, and black are also common. It is generally believed that the colour is a function of particle size, chemical purity, stoichiometry, and method of preparation although the colour of PuO2 resulting from a given method of preparation is not always reproducible [16].
The concentrations of impurities in the Pu batch n o 716 measured by ICP-MS are presented in Table 1 and expressed in μg/g of the original solid material. As the dissolution was not complete, results should be considered as indicative. Other lanthanides could also be seen at trace level (few ppm). The actinide elements come from the radioactive decay of plutonium. The Ce could be related to the tests with CeO2 in the glove box.
The average concentrations of impurities in Pu batch n o 716 showed also V, which was confirmed by SEM-EDX. The EDX analyses showed the presence of several elements in Pu batch n o 716. Table 2 presents the weight percentages in the powder at positions 12-13-14-15 of the SEM image (Fig. 3). The SEM image showed a morphology of irregular crystals, rosettes, and thin plates of different sizes. A review of the literature suggested that the morphology of PuO2 formed through the Pu(IV) oxalate process can differ based on the parameters and conditions chosen during sample preparation [17,18]. The presence of C could be linked to a thermal decomposition process of the hydrated Pu oxalate Pu(C2O4)2·6H2O to produce PuO2 powder. This process results in the oxalate ligand decomposition and production of amorphous Pu oxycarbide species, CO2, and ultimately crystalline PuO2 with some residual Ccontaining species [17]. This would also explain the colour of the powder, which was black and light brown. If the decomposition would have been successful, the expected colour of the PuO2 powder would have been homogeneous black [18]. The XRD analysis showed that the sample was not homogeneous and pure. It contained at least 2 phases a c b with PuO2 as the main component. The peak shape, broad at bottom and sharp at the top relates to the heterogeneity in the crystallinity of the sample. This indicated an amorphous part. Based on a phase search in the ICSD and PDF2 databases using the elements and element family of Pu, O, C, Cl and Np as reported in the ICP-MS and SEM-EDX, the second phase could be assigned to a compound isostructural to H2F12Np3O [98-004-7190] with a monoclinic structure. However, a compound isostructural to PuF4(H2O)1.6 [01-074-1766] with a cubic structure could not be excluded. This second phase, present at about 15 wt% in the sample, included a halogen element. F was given because this element was known in the database of phases. It could also be Cl instead of F as PuCl4 has the same structure as PuF4 and Cl was also confirmed by EDX analyses.
The DTA-GTA displayed about 11 wt% weight loss for PuO2 powder in an O2 and in an Ar atmosphere. Along the temperature, the weight decreased regularly till about 700 °C and then along two steps, [700-800] °C and [800-1100] °C. The pellet showed a weight loss of 9% under O2, while for Ar the final step was not reached yet at 1400 °C. As expected, the reaction was stronger with the powder than with the pellet due to the larger surface available for the reaction. The weight loss took place along three steps: 1) <700 °C; 2) >700 °C and >800 °C; and 3) >800 °C. The relatively important weight loss of 11 wt% as a result of the DTA-GTA on the Pu batch n o 716 showed that the sample could not be pure PuO2. By comparison, for pure Pu-oxalate, a weight loss of about 35 wt% is expected.

Conclusion
Several pellets of stable Mo metal powder and of 239 PuO2 powder were prepared as targets for neutroninduced cross section measurements within SANDA [11] and EUFRAT [14], respectively. The pellets were produced by uniaxial pressing the powders with a hydraulic press, without adding fluids, additives or lubricants. The pressure force and the diameter of the pellets were optimized to produce mechanically stable unsintered pellets with a mass of 2 g and a thickness as thin as possible and maximum 1 mm. The PuO2 pellets made of PuO2 of batch n o 716 were characterised by ICP-MS, SEM-EDX, XRD and DTA-GTA.
The Mo pellets had good strength, sharp borders and glossy surfaces and could be further handled without loss of material. The compaction of the PuO2 powder however resulted in bad strength quality for the purified batches. Only pellets from the Pu batch unpurified for Am could be produced with a good strength.
Powder contamination can influence the pressing process. The amount and the type of impurities depend on the powder production process, the storage container, the purification process, the process environment, etc., which is not always known. Crystalline order and morphology of PuO2 grains and aggregates are known to be sensitive to the initial production conditions [19]. The analyses of the PuO2 of batch n o 716 by ICP-MS, SEM-EDX, XRD and DTA-GTA showed beside the expected Pu, O and actinide elements from the radioactive decay of Pu, the important presence of V, C, and Cl. C could be linked to the purification process of PuO2 via thermal decomposition of hydrated Pu(IV) oxalate. We can only assume that the presence of C, CO2, and Cl in the purified Pu, in combination with water absorbed from the environment, caused corrosion of the steel press tools. This could result in end-capping by friction of the Pu with the inner wall of the die sleeve and/or that volatile impurities caused high porosity and cracks in the Pu pellet. A heat treatment of the Pu material under oxidizing atmosphere would favour the purification of the material. It would desorb and decompose impurities formed by the moisture, potential presence of chloride and rest of oxalate if any.