Exploring a surrogate of Pellet–Cladding interaction: Characterization and oxidation behavior

The present work analyzes the effect of different Zr contents on the microstructure and thermal behavior of non-irradiated UO2. A set of Zr-doped UO2 (0, 20, 40, 80, and 100 wt%) pellets was prepared via solid-state synthesis by mimicking the chemical bonding between ZrO2 and UO2. After sintering, the Zr-doped UO2 monoliths were characterized by scanning electron microscopy, Raman spectroscopy, and X-ray diffraction. Oxidation of the Zr-doped UO2 samples has been monitored by thermogravimetric analysis. Results of thermal behavior under air and low oxygen partial pressure show a profound effect of delayed oxidation with the addition of Zr to UO2 and then an increased oxidation resistance of UO2 to U3O8 when compared to pure UO2 pellet.


Introduction
In light water reactors (LWRs), UO 2 pellets are mainly used as fuel within a zirconium-based alloy clad to protect the fuel pellets and to avoid radionuclides releasing into the coolant [1]. During in-reactor operation radiation effects, thermal expansion and microstructural alterations, i.e., the production of gaseous elements (mainly Xe and Kr) and Fission Products (FP), induce dimensional changes and swelling of the fuel pellet [2,3]. Particularly, the thermal gradient at which the nuclear fuel is subjected during operation is one of the reasons of the deformation caused in the pellet, which could affect to its interaction with the cladding. Thus, a potential pellet-cladding contact could take place because of a decrease in the cladding diameter (creep-down due to pressure from the primary coolant) [2] and an increase in the pellet diameter. This phenomenon is known as Pellet-Cladding Interaction (PCI) and occurs in the majority of commercial fuel rods [3].
The contact between pellet and cladding, given through a chemical bonding layer, first occurs at the inter-pellet spaces while a continuing pellet fragmentation (radial and axial cracks) takes place simultaneously. The interaction layer consists of two regions. First region occurs closer to the cladding (polycrystalline ZrO 2 ) and the second layer nearer the fuel pellet is an interphase formed by solid solutions of (U, Zr)O 2 and an amorphous phase (induced by radiation). This second layer is composed by variable relative concentrations of U and Zr [2]. Because it may induce cladding fuel failures in commercial LWRs during a power transient, it is considered as a mid-priority phenomenon in the investigation of the safety approach under dry storage [4]. In addition, it is one of the causes identified as leading to potential fuel failure. Most of the related literature with UO 2 -ZrO 2 solid solution system comes from severe accident scenarios such as Fukushima in Japan [5,6]; because of core melting in Nuclear Power Plants (NPPs), U and Zr could interact.
Moreover, the risk of enhanced PCI grows with burnup. In fact, evidences on Zr-oxide layer formation and Zircaloy-Fuel gap closure were observed in high-burnup fuels (HBU). Studies conducted on fuel with burnups > 55 MWd·kg U −1 showed the formation of a restructured region, named as "rim" structure or HBU structure at the periphery of UO 2 pellets [7]. As the burnup increases, some effects turn out to appear in this new region, i.e., smaller grains, lattice contraction (recrystallization), higher porosity, hardness decrease, and the closure of pellet-cladding gap in fuel rods [8].
Understanding the potential chemical oxidation resistance of UO 2 (matrix fuel) to U 3 O 8 by ZrO 2 /Zr system in case of undetected damaged cladding (zirconium alloy), and a potential air intrusion, is relevant in terms of assessing fuel integrity. Therefore, assuming evidences on Zircaloy-Fuel gap closure (Zr-UO 2 layer formation from HBU PWR fuels) [9][10][11], here we conduct characterization of Zr-doped UO 2 pellets including morphological, chemical, and crystallographic analyses. Furthermore, we present some results based on the oxidation behavior of these samples under dry or transportation conditions that probe the chemical oxidation resistance of UO 2 to U 3 O 8 , when they form part of the (U, Zr)O 2 solution, and then mimicking the chemical bonding between ZrO 2 and UO 2 .
Mixtures of Zr and U oxides (including undoped UO 2 and ZrO 2 ) were prepared by mechanical dry blending from mixing weighted amounts of UO 2 and ZrO 2 powders, together with 1 wt% of binder EBS (Ethyl Bis Stearamide, C 38 H 76 N 2 O 2 , Tokyo Chemical Industry) [16], using low-energy ball milling. Then, powder mixtures were sintered at 1675 °C for 4 h in a reducing atmosphere flow (N 2 -4.7%H 2 ) [17,18]. Samples were characterized after being sintered: surface morphology and average grain size calculation by SEM; BET SSA (Specific Surface Area) with N 2 ; the purity and crystalline structure by both XRD and RQPA (Rietveld Quantitative Phase Analysis); Raman spectroscopy (laser λ exc = 633 nm) and density by Archimedean immersion.

Characterization of mixed oxides sintered pellets
In the recorded XRD pattern of pure UO 2 , the presence of a single cubic phase confirms the purity of the sample, and the (111) peak at low 2θ ~ 28.558° is consistent with the presence of a slightly hyperstoichiometric layer UO 2+x on the surface and with a thickness of the order of tens nm [19]. In addition, the diffractograms at doping concentrations of 20 and 40 wt% ZrO 2 clearly reveal the apparition of an additional diffraction peak at 2θ ~ 29.8° due to the tetragonal ZrO 2 phase, which decreases in the sample doped with 80 wt% ZrO 2 , once monoclinic lattice appeared, and completely fades in pure ZrO 2 . This observed peak is in agreement with literature data [15,20]. Indeed, m-ZrO 2 is the thermodynamically stable phase at room conditions [15]. Also, in this sample, an additional peak growth on 2θ ~ 31.5° suggests a non-dissolved fraction of ZrO 2 in the UO 2 matrix upon sintering or cooling [20]. From the results by RQPA, one might think that a solid solution should be also present in the tetragonal and the monoclinic phase. Anyhow, the solubility of UO 2 seems to be greater in the tetragonal phase since it shows higher differences in the lattice parameters than that from the pure ZrO 2 phase. All these facts mean that, from the data obtained by XRD, it is possible to get an approach on the solubility limit of Zr in the cubic UO 2 matrix in these conditions. The UO 2 solid solution was cubic and extended up to 20 wt% ZrO 2 . Segregation of tetragonal ZrO 2 phase is identified even at concentrations below 40 wt% of ZrO 2 . Far beyond this ZrO 2 concentration, the monoclinic phase starts to appear. Figure 3 shows the microstructure of the polished sintered pellets in backscattered electron (BSE) mode. Normalized Raman spectra recorded on sintered pellets are displayed in Fig. 2B. Pure UO 2 spectrum is characterized by the presence of the same features than that described in Fig. 1. The weak bands observed at ~ 570 cm −1 and at ~ 630 cm −1 are assigned to the LO mode (lattice distortion), i.e., structural defects in the cuboctahedral symmetry of the interstitial oxygens, respectively, due to slight oxidation to UO 2+x . This is consistent with XRD results, confirming the formation of an  The overall interpretation from all these analysis reveals non-uniform distribution of Zr in the UO 2 matrix and ZrO 2 segregation in grain boundaries, presumably because the solubility limit is reached in the fabrication procedure.

Oxidation behavior of mixed oxides (U, Zr)O 2
Using TGA in air, the thermal oxidation behaviors of (U, Zr)O 2 samples containing a range of Zr concentrations are assessed. To understand the effect of Zr on the formation of U 3 O 8 , the in situ analysis of the oxidation reaction performed by TGA at different oxygen partial pressure conditions is presented in Fig. 4 for all the studied materials.
Differences in oxidation behavior of the studied (U, Zr) O 2 pellets as a function of Zr content (0-100%) and oxygen partial pressure are observed. In general, a delay in the UO 2 oxidation is found when increasing the Zr doping concentration. XRD analyses of the oxidized sample (not shown) from pure UO 2 confirmed complete conversion to U 3 O 8 . When samples were oxidized in air, the two representative oxidation transitions (UO 2 → U 4 O 9 /U 3 O 7 → U 3 O 8 ) are observed as a two-step weigh change. However, this behavior is not followed for the sample doped with 80 wt% ZrO 2 . On the contrary, in 1%O 2 -N 2 , the two-step transitions are continuously softened with greater Zr content.
The TGA indicates that, as expected, increasing Zr content on UO 2 matrix enhances the stability of the intermediate U oxidation products and slowed the subsequent oxidation rate.

Conclusions
We examined the effects of Zr doping on the thermal stability of UO 2 under dry interim storage conditions (air or 1%O 2 -N 2 ). Mixtures of UO 2 and ZrO 2 in a range of possible

B) N 2 -1%O 2°
compositions across the layer (PCI phenomenon) are tested to examine the effect of Zr content and oxygen concentration (gas phase) and simulate ZrO 2 /UO 2 bonding (chemical adhesion) in HBU fuels. Systematic oxidation studies on x wt%ZrO 2 -UO 2 (x = 0, 20, 40, 80, and 100) solid samples were analyzed under non-isothermal thermoanalytical conditions (10 ºC min −1 , 900 ºC). The thermal analysis indicates hindered of matrix oxidation attributed to the presence of zirconia. The present preliminary results studied as a representative example of a typical HBU fuels and PCI showed that the presence of Zr in UO 2 fuel matrix provides oxidation resistance to U 3 O 8 .
Therefore, the results of this analysis are important for understanding the behavior of SNF under dry or transportation conditions, once closure of pellet-cladding gap occurs. In order to improve this assumption to real irradiated fuel, further research involving other characterization techniques and considering the occurrence of FP in the Zr/ZrO 2 must be explored.