Unlike conventional sintering, the exact SPS processing conditions required .... driver for the adoption of SPS for MOX fuels than UO2 on account of their higher ...
Project acronym: NUGENIA-PLUS Grant agreement no: 604965
Deliverable: D6.35 SPARK plasma sintering research in nuclear technology (public summary)
Version/date: V1/29-09-2016
Author(s): D. Shepherd, D. Goddard (NNL); K. Johnson (KTH); M. Cologna (JRC-ITU)
The research leading to these results is partly funded by the European Atomic Energy Community’s (Euratom) Seventh Framework Programme FP7/2007-2013 under grant agreement No. 604965
1
Content 1.
Introduction ..................................................................................................................................... 3
2.
UO2 fuels .......................................................................................................................................... 3 2.1.
Assessment of the O/M ratio after SPS ................................................................................... 4
2.2.
Annealing tests in Ar/H2 and CO/CO2 mixtures to remove the C layer ................................... 5
3.
MOX fuels ........................................................................................................................................ 6
4.
Inert matrix fuel wasteforms ........................................................................................................... 7 4.1.
Assessment of homogeneity ................................................................................................... 7
4.2.
Assessment of oxidation behaviour in air ............................................................................... 9
5.
Wider nuclear applications of SPS ................................................................................................... 9
6.
Conclusions.................................................................................................................................... 10
7.
Future Activities............................................................................................................................. 10
8.
References ..................................................................................................................................... 11
2
1. Introduction Spark plasma sintering (SPS) is a beyond state-of-the-art fabrication technique that has become available for metal and ceramic powder processing. It works by passing a powerful current through a die and powder compact during pressing, with graphite being the most commonly used die material. SPS rapidly sinters the compact within a few minutes, potentially allowing a net shape component to be produced without further time-consuming and wasteful machining. In addition, the as-produced microstructure can be highly refined (small grain size), which gives associated material property benefits such as increased mechanical strength and toughness. As a result, SPS has been identified in the nuclear sector as potential fabrication technology for a number of different components. One of the main foreseen nuclear applications of SPS is fuel pellet manufacturing. If applied in a conventional fuel manufacturing line, then through its net shape or near net shape forming ability, SPS has the potential to eliminate or reduce the need for grinding to achieve the final geometry. Grinding generates hazardous dust residues, which are especially problematic for the production of high radioactivity fuels such as those containing plutonium and/or minor actinides. SPS would also combine the pressing and sintering stages, which could save time and reduce energy consumption. The FP7 NUGENIA+ SPARK project has investigated experimentally some of the barriers to the adoption of SPS in the area of fuel manufacturing in order to progress its deployment as a safe and economic processing method. A literature review of wider nuclear applications of SPS has also been performed.
2. UO2 fuels Contributors : M. Cologna, C. Boshoven, D. Bouexiere, M. Ernstberger, D. Freis, H. Hein, M. Holzhäuser, V. Tyrpekl, J. Somers ; Joint Research Centre – Institute for Transuranium Elements (JRCITU) Oxide fuels, primarily UO2, are used in the vast majority of nuclear reactors. A known issue of SPS is the formation of a thin reacted layer on the pellet surface, due to contact with the graphite dies. Such a layer may be removed by grinding but this not desirable in terms of realising net shape manufacturing. Furthermore, the stoichiometry of oxide fuel pellets (oxygen to metal ratio) is of paramount importance, since it governs important properties that determine the safety of the fuel such as thermal conductivity, mechanical and diffusional behaviour. The stoichiometry has thus to be strictly controlled during synthesis. Unlike conventional sintering, the exact SPS processing conditions required for adjustment of the O:M ratio were not completely known. This task has investigated possible short post-thermal treatments of SPS-manufactured UO2 pellets to remove the reaction layer and to adjust the O:U ratio to meet a representative fuel specification at the same time. A series of UO2 pellets were prepared by SPS and annealed in different well-defined oxygen partial pressures in a conventional furnace and in a thermobalance. A commercial UO2 powder obtained from the ammonium diuranate (ADU) route was used, with an O:U of 2.16. Sintering was conducted in a FCT Systeme SPS in a glovebox for handling radioactive substances. Both glovebox and SPS chamber were kept under Ar atmosphere. The powder was loaded into 6 mm diameter graphite dies, without using any graphite separation paper. About 300 mg of powder was used each time. The heating and cooling rates were 200°C/min and a uniaxial 3
pressure of 70 MPa was applied constantly from room temperature. The maximum temperatures and dwell times were varied. The microstructures of the sintered samples were examined using scanning electron microscopy (SEM) whilst the phase composition and O:U ratio was measured by XRay Diffraction (XRD) using the Rietveld methodology.
2.1. Assessment of the O/M ratio after SPS The conditions for the control of O:U are well known for conventional processing; one typical route to obtain dense UO2.00 pellets involves sintering hyperstoichiometric (UO2+x) powder in Ar/H2 for several hours at 1700°C. For SPS of UO2+x, the suitable sintering and reduction conditions are more complex to identify. In particular, the oxygen potential of the gas in the die cavity is not well defined and can differ substantially from the external set gas atmosphere. This is because above 600°C, the reaction of the graphite die with oxygen from the sample creates CO and thus a reducing atmosphere, whose oxygen potential is hard to evaluate. The external field and high current passing through the die/powder compact assembly may also play a role, which is not yet fully understood. Ge et al. [1] studied the evolution of O:U of UO2+x in SPS and found that the stoichiometry decreased monotonically with the maximum hold time and temperature. These findings were explained by a solid state reaction of the excess oxygen in UO2+x with the graphite of the tools. In this study, the stoichiometry of SPS UO2 pellets sintered under different conditions is summarised in Figure 1. It can be seen that O:M decreases with maximum temperature and hold time; over 1hr at 1000°C is needed to reach UO2.00 compared to only 30s at 1450°C.
Starting powder 2.16
2.15 600°C
750°C
2.10 O/M
800°C
2.05 1050°C
850°C 1000°C
2.00
>1450°C
0
UO2.00
1350°C
10
20
30
40
50
60
Hold time (min)
Figure 1:
Average O:U of SPS UO2 pellets vs. holding time and maximum temperature. Hollow data points are from Ge et al. [1]
Therefore with SPS, the conditions can be sufficiently reducing to reach UO2, even in the absence of a specific processing gas unlike conventional sintering. In addition, densities above 95% can be achieved at temperature as low as 1000°C in a few minutes by SPS. However, this temperature is not sufficient to achieve UO2.00 within 1hr holding time, and thus a post treatment in a reducing 4
atmosphere (e.g. Ar/H2) would be needed to adjust stoichiometry. Depending on the microstructure that is sought, alternative solutions are the use of higher temperatures, longer holding times, or the use of pre-reduced UO2 powder as in [3]. In order to investigate the variation of stoichiometry within the pellet and its relation to the direction of the electric field, O:U of the top (anode) and bottom (cathode) end faces of two low temperature SPS pellets (800°C and 1000°C for 5 minutes) were measured and are recorded in Table 1. The lower oxidation state of the surfaces compared to the average (centre) can be explained by reduction by graphite surrounding the pellet forming CO. The higher O:U remaining at the anode was most likely due to the effect of the electric field, which acts to drive the highly mobile negatively charged oxygen anion towards the anode. These two factors will give a complex overall distribution of O:U. Table 1:
Stoichiometry of the end faces of SPS UO2 pellets
Sample
800°C anode
800°C cathode
1000°C anode
1000°C cathode
O/U
2.14
2.03
2.05
2.00
2.2. Annealing tests in Ar/H2 and CO/CO2 mixtures to remove the C layer Since the typical processing dies for SPS are made of graphite, the surfaces of the pellets are inevitably contaminated with carbon. The formation of a thin contact layer consisting of graphite and uranium carbides (UCx) was observed for samples treated at 1450°C [1]. At processing temperatures below 1200°C, the contact layer will consist predominantly of graphite [2, 3]. In this study, the effectiveness of post-SPS annealing to remove the contact layer was studied, both in Ar/H2 as well as CO/CO2. The first tests were done by annealing in Ar/H2 in a conventional furnace. For such tests ThO2 was used instead of UO2 in order to rapidly confirm the success of the heat treatment; the colour of ThO 2 is white, while the colour of the C contact layer is black, and its removal on ThO2 is evident at the end of the annealing without the need of more time consuming analysis techniques. Images of the pellets are shown in Figure 2.
Figure 2:
SPS ThO2 before and after Ar/H2 annealing at different temperature
The samples treated at a temperature of 1375°C and above showed an upper white surface, while the upper surface of the sample annealed at 1060°C was still black. The lower surfaces in contact with the molybdenum tray did not turn white in any condition, which could be due to insufficient exposure to the Ar/H2 flow. However the darker appearance of these surfaces did not disappear even 5
after a second heat treatment done under the same conditions with the samples turned on their side. Therefore although tests on ThO2 show that annealing in Ar/H2 at 1375°C is sufficient to remove the C layer from all the free surfaces, pellets should be placed on a tray that minimises the contact of the sample with the tray itself, for example a porous molybdenum foam or a bed of spheres in order to assure the full removal of carbon. In the case of UO2+x, such treatment in Ar/H2 would also have the advantage of ensuring reduction to UO2.00. Annealing in 1:100 CO:CO2 at 1200°C for 2.5hr was also found to be effective at removing the contact layer. However the final stoichiometry of the pellet under such conditions was higher than UO2.01 and so a two-step process would be needed to reduce to UO2.00, by switching the atmosphere to a more reducing gas (e.g. CO/CO2 with higher CO content or Ar/H2).
3. MOX fuels Contributors : D. Goddard (National Nuclear Laboratory), J. Turner (University of Manchester). MOX (mixed U-Pu oxide) fuels are used commercially in many reactors. There may be a stronger driver for the adoption of SPS for MOX fuels than UO2 on account of their higher radioactivity and increased hazard from grinding dust, which the net shape potential of SPS could mitigate. The microstructure of any SPS MOX fuel must still meet commercial specifications. In particular, homogeneity is a key requirement, as large Pu rich regions (>400µm) are associated with the accumulation of solid and gaseous fission products during irradiation, resulting in localised swelling. It should be noted that the different milling techniques practised in different countries do produce fuels with varying degrees of homogeneity. In this task, SPS MOX fuel was fabricated using Ce as the most commonly used surrogate for Pu in order to save dose, time and cost during this preliminary study. It is expected that the results will be readily transferrable to genuine MOX fuel. Figure 3 shows a (U,Ce)O2 disk containing 7mol%Ce. This was produced using 20 mm diameter graphite dies and foil in a Thermal Technology LLC SPS Model 10-4 using 40 MPa and a dwell time of 5 minutes at 1400°C. The peak shift in the XRD spectrum in Figure 3 was consistent with the formation of a solid solution and similar to that observed for a conventionally processed (U,Ce)O2 pellet.
Figure 3:
Image of (U0.93,Ce0.07)O2 pellet and XRD indicating the formation of a solid solution 6
4. Inert matrix fuel wasteforms Contributors: K. Johnson, D.A. Lopes, J. Wallenius ; Kungliga Tekniska Högskolan (KTH) Inert matrix fuels (IMFs) aim to destroy fissionable and/or transmutable materials within a nonfissile/fertile matrix in order to minimise residual fissile material and long lived radiotoxicity within a form that is suitable for direct waste disposal without further conversion. The greatest interest is for the destruction of plutonium and of the long-lived and highly radioactive minor actinides. A possible additional benefit of SPS compared to conventional sintering for the processing of such IMFs, lies in its potential to suppress the volatilisation of elements such as americium. KTH has been developing ZrN matrix IMF for a number of years and targets the use of SPS for the manufacture of such fuels, which have been shown to be difficult to produce with a density greater than 95% of theoretical using conventional methods [4, 5]. Nitride fuels are of interest due to their potential higher density and thermal conductivity compared with oxide fuels, which may be advantageous with respect to safety and economics. SPS (Zr,U)N has been produced using U as surrogate for Pu and Pu/Am. Earlier work [6] has shown a tendency for these fuels to form an incomplete solid solution (
