Gd₂O₃ Doped UO₂(s) Corrosion in the Presence of Silicate and Calcium under Alkaline Conditions

Abstract

The anodic reactivity of UO₂ and UO₂ doped with Gd₂O₃ was investigated by electrochemical methods in slightly alkaline conditions in the presence of silicate and calcium. At the end of the experiments, the electrodes were analysed by X-ray photoelectron spectroscopy to determine the oxidation state of the uranium on the surface. The experiments showed that the increase in gadolinia doping level led to a reduction in the reactivity of UO₂, this effect being more marked at the highest doping level studied (10 wt.% Gd₂O₃). This behaviour could be attributed to the formation of dopant-vacancy clusters (Gd^III-Ov), which could limit the accommodation of excess O²⁻ into the UO₂ lattice. In addition, the presence of Ca²⁺ and SiO₃²⁻ decreased the anodic dissolution of UO₂. In summary, the Gd₂O₃ doping in presence of silicate and calcium was found to strongly decrease the oxidative dissolution of UO₂, which is a beneficial situation regarding the long-term management of spent nuclear fuel in a repository.

1. Introduction

The repository concept is an internationally accepted approach for the long-term management of spent nuclear fuel, where the fuel would be sealed in containers. However, in the worst-case-scenario where the containers fail, the contact of the fuel with groundwater could be possible. Since the majority of the radionuclides are located within the UO₂ matrix and taking into account that U^VI is more soluble than U^IV, the release of the radionuclides to the groundwater would be controlled by the fuel corrosion/dissolution rate, which would be strongly affected by the redox conditions. Under oxidizing aqueous conditions, the corrosion of UO₂ occurs via two steps process. First U^IVO₂ is oxidized to U^IV_1−2xU^V_2xO_2+x forming a thin oxidized surface layer, while in the second step soluble U^VI is released to the solution [1]. In alkaline conditions, dissolved uranium could reprecipitate, forming corrosion product deposits (UO₃·yH₂O).

Some of the main parameters affecting the corrosion of the UO₂ matrix in a repository are the rare earth (RE) doping of the fuel [2,3] and the composition of the groundwater [4]. Nowadays, UO₂ is commonly doped with rare earth elements such as Gd (in the form of Gd₂O₃) [5] for its critical role as a burnable neutron absorber. The high neutron absorption cross section of ¹⁵⁵Gd and ¹⁵⁷Gd helps to counterbalance the excess of reactivity in the reactor core during the initial stages of operation, improving overall reactor performance.

The influence of the incorporation of Gd to UO₂ has been extensively studied to determine lattice parameters [6,7], UO₂ oxidation in dry experiments [8,9], UO₂ electrochemical reactivity [1,7,10] and UO₂ dissolution [11,12].

Air oxidation kinetics of UO₂ at high temperature (325 °C) indicated that doping with Gd stabilizes the first oxidation product and inhibits the oxidation to U₃O₈ being the kinetic mechanism dependent on the Gd content [8]. He et al. studied the effect of fission product doping on the anodic reactivity of UO₂ (they worked with SIMFUEL) and concluded that doping with such elements distorts the lattice structure of UO₂. This fact affects its corrosion process, since the oxidation was suppressed as the extent of doping increased [1]. Liu et al. showed that Gd doping lead to a contraction of the fluorite lattice, reducing the electrochemical reactivity of (U_1−yGd_y)O₂ [7]. Razdan and Shoesmith studied the influence of trivalent-dopants (such as Gd^III) on the electrochemical behaviour of UO₂ [2]. They found that while the mechanism of oxidation/dissolution of Gd-doped UO₂ is comparable to that of SIMFUEL, the reactivity is lower. The effect of Gd-doping was also observed on dissolution experiments, showing a reduction of the oxidative dissolution rate [11,12,13]. This behaviour was commonly attributed to the formation of Gd^III-Ov clusters [14] in order to maintain electroneutrality, limiting the accommodation of O²⁻ ions during oxidation [15]. Therefore, the Gd^III influence on fuel reactivity under repository conditions is of great importance.

Concrete has been proposed as a construction material in some deep geological repositories [16]. Water in contact with concrete presents high pH values and contains silicate and calcium ions [17]. Flow-through dissolution experiments with UO₂ performed by Wilson and Gray [4] revealed a considerable reduction in the concentration of U measured when Ca and Si were present in the solution. Santos et al. investigated the effect of SiO₃²⁻ [18] and Ca²⁺ on the corrosion of SIMFUEL [19]. The authors concluded that these ions had little influence on the first oxidation step, but they inhibited/retarded the formation of U^VI, probably due to the adsorption of silicate and calcium on the fuel surface. Espriu et al. studied the corrosion of SIMFUEL at high pH and in the presence of calcium and silicate [20]. Their X-ray photoelectron spectroscopy (XPS) analysis evidenced that the presence of both ions in a high alkaline solution have a significant impact on the oxidation of the surface of SIMFUEL, since the surface was less oxidized than that of the experiments conducted in a solution without such ions. However, the corrosion of Gd₂O₃ doped UO₂ under cementitious water is not well known.

In this work, both undoped UO₂ and UO₂ doped with 5 and 10 wt.% Gd₂O₃ pellets were examined electrochemically in order to study the influence of Gd₂O₃ doping on the oxidative reactivity of UO₂ at slightly alkaline groundwater in the presence of silicate and calcium.

2. Results and Discussion

2.1. Characterisation of the Gd₂O₃ Doped UO₂ Pellets

The surface morphology of the sintered UO₂ pellets doped with 5 wt.% and 10 wt.% Gd₂O₃ (Figure 1), exhibited clearly defined grain boundaries and low-porosity microstructure. Furthermore, no unreacted Gd₂O₃ was found and EDX analysis revealed a homogeneous distribution of gadolinium within the UO₂ matrix.

The examination of the samples using X-ray diffraction (XRD) revealed the fluorite crystal structure of UO₂ (Figure 2). The peaks observed in the Gd₂O₃-doped samples were found to be slightly shifted toward greater angles compared to the UO₂ sample. This shift was caused by the lattice contraction as gadolinium concentration increased, as previously observed in previous studies [7].

2.2. Cyclic Voltammetry

Figure 3 shows a series of cyclic voltammograms (CVs) recorded in silicate and calcium solution on the undoped and Gd₂O₃ doped UO₂ electrodes. Similar stages of both oxidation and reduction regions were previously observed for SIMFUEL [21] and undoped UO₂ [22]. In the forward scan, two different regions can be observed. In region 1 (at the potential of around −0.2 V), the increase in the anodic oxidation current observed is attributed to the formation of a mixed U^IV/U^V thin surface oxide layer with a thickness limited by the O²⁻ injection into the UO₂ fluorite lattice. It can be observed that the onset of the oxidation in region 1 is independent of the Gd₂O₃ doping level. However, there is a significant difference in the anodic current observed, registering lower current density values as the doping level of Gd₂O₃ increases. This behaviour was observed in [10] for hyperstoichiometric U_1−yGd_yO_2+x and can be attributed to the ability of Gd to retard the oxidation of UO₂ matrix surface by the formation of Gd^III-Ov clusters. In region 2 (>0.2 V), the thin surface layer is further oxidized to soluble U^VI and then the dissolution as UO₂²⁺ could take place. Under alkaline conditions, the solubility of UO₂²⁺ is limited [1], therefore it could be deposited as insoluble U^VIO₃·yH₂O on the electrode surface. The formation of this deposit could explain the suppression of the current observed in Figure 3 at higher voltages (>0.3 V). Another possibility is that the silicate and calcium in solution could be adsorbed on the electrode surface, suppressing the further oxidation of the thin surface layer [19]. The stabilization of the UO₂ matrix in presence of silicate and calcium was previously observed on SIMFUEL [18,19,20].

On the reverse scan, a cathodic reduction peak (region 3) was observed with a maximum at around −0.7 V. This reduction peak is commonly attributed to the reduction of U^IV_1−2x/U^V_2xO_2+x /U^VIO₃·yH₂O surface layer [2]. The presence of this reduction peak for all electrodes indicate that an oxidized surface layer is formed despite the presence of Gd, however, the current associated with the reduction peak vary significantly with doping level. This confirms that the thickness of the thin U^IV_1−2x/U^V_2xO_2+x layer differs among them, which is in agreement with the current densities registered in regions 1 and 2.

In contrast, CV experiments performed on 5 wt.% Gd₂O₃-doped UO₂ electrode in presence of bicarbonate (Figure 4) registered higher anodic oxidation currents in regions 1 and 2 than in solutions that contained silicate and calcium. This behaviour can be attributed to the ability of bicarbonate to form soluble uranyl carbonate complex ions, which avoided the formation of secondary phases on the electrode surface and enhanced the oxidation and dissolution process of UO₂ [23]. This is consistent with the smaller reduction peak observed in region 3 in bicarbonate solution.

The suppression of the anodic dissolution could be probably attributed to the adsorption of silicate and calcium to the electrode surface. These results are in good agreement with the findings of Wilson and Gray [4], as they detected in leaching experiments a thin layer of Ca and Si, which inhibited the dissolution of UO₂. Further investigation in solutions containing Ca, Si and carbonates, which better mimic cementitious water, may be needed to determine whether the predominant factor is the formation of soluble ternary complexes Ca-U(VI)-CO₃ [24] or the inhibitory effects of Ca and Si.

2.3. Potentiostatic Oxidation

Since CVs are rapid experiments, to better observe differences in reactivity due to doping level, potentiostatic oxidation experiments were performed. Figure 5 shows a series of current density vs. time plots recorded at an oxidative potential of 0.4 V for 1 h for the different electrodes in a solution with silicate and calcium. In general, for all electrodes, the anodic current decreases linearly on the logarithm scale with time, which is consistent with a loss of surface reactivity due to the growth of a U^IV/VO_2+x surface layer [19].

For the non-doped electrode, the current density registered was considerably higher than in doped electrodes. After 1000 s the current tends toward a steady-state value (reaches a plateau). Regarding doped electrodes, 5 wt.% Gd₂O₃-doped UO₂ electrode registered at short times higher current densities compared to 10 wt.% Gd₂O₃-doped UO₂ electrode, while at long times, the currents continued decreasing, reaching very low values for both electrodes.

Figure 6 shows the cumulative integration of the potentiostatic current densities (from Figure 5) as a function of time for the three electrodes. The general trend observed is that the charge accumulated increases as the doping level decreases. Whereas the total charge for the non-doped electrode continues increasing with time in the period of time registered (3600 s), for doped electrodes their respective total charges tend to reach a stable state.

XPS was performed after potentiostatic experiments to determine the composition of the electrode surface as a function of the extent of doping. Figure 7 shows the deconvoluted spectra for the U 4f_5/2 and U 4f_7/2 regions and their corresponding satellite structures for the three different electrodes while

Table 1. Relative fractions of U^IV, U^V and U^VI on the surface of the different samples from XPS results after potentiostatic oxidation at 0.4 V.

Sample	UIV(%)	UV(%)	UVI(%)
UO2	8 ± 5	26 ± 5	66 ± 5
UO2—5 wt.% Gd2O3	7 ± 5	35 ± 5	58 ± 5
UO2—10 wt.% Gd2O3	15 ± 5	62 ± 5	23 ± 5

collects the percentages of U^IV, U^V and U^VI on the electrodes surface. In all cases, as expected, traces of U^IV were detected, whereas U^V and U^VI states dominate the surface composition of the electrodes, since an oxidizing potential was applied. This is consistent with the formation of an underlying U^IV_1−2xU^V_2x O_2+x layer and the following formation of UO₂²⁺, which can be accumulated as a U^VIO₃·yH₂O deposit. It can be observed that as the doping level increases, the U^V content increases, at the expense of U^VI. Although the effect of doping with 5 wt.% Gd₂O₃ is small, in the 10 wt.% Gd₂O₃-UO₂ electrode, a strong U^V signal was obtained, which is consistent with a marked decrease in reactivity at high doping levels (10%), as already reported in the literature [7]. The differences observed in U^VI concentration provide evidence that doping with Gd₂O₃ slows down the second oxidation step (U^IV_1−2xU^V_2x O_2+x to U^VIO₃·yH₂O). These results are in agreement with the potentiostatic data, since lower anodic charges lead to lower amount of U^VI and hence, lower oxidation.

This observation is consistent with the results reported in air oxidation by Scheele et al. [25], who observed a delay on the onset of the second oxidation to U₃O₈ on Gd₂O₃-UO₂ doped samples. XPS analysis performed after dissolution experiments in presence of carbonate/bicarbonate [12] revealed a big contribution of U(V) on the surface of the Gd₂O₃-UO₂ samples, and low U concentration on the leaching, indicating that Gd-doping enhanced the stability of U(V) and inhibited the subsequent oxidation to U(VI). A decrease in the uranium concentration on Gd-doped UO₂ pellets was also reported in leaching experiments in H₂O₂ and γ-irradiation [13]. Similar behaviour was found in the literature for Nd-doped UO₂ in presence of H₂O₂ [26]. This found could indicate that doping UO₂ with trivalent dopants (like Gd³⁺ and Nd³⁺) would lead to a comparable stabilization effect on the UO₂ structure.

2.4. Cathodic Stripping Voltammetry

Additional potentiostatic experiments were performed at the same conditions (0.4 V for 1 h), but this time followed by cathodic stripping voltammetry (CSV) in order to reduce the surface layer and the deposit. Figure 8 shows the CSV recorded for the three electrodes. Two reduction peaks were observed in this region, which are commonly assigned to the reduction of the dual phase nature of U^IV_1−2xU^V_2x O_2+x/U^VIO₃·yH₂O. The reduction peak observed at −0.7 V (peak 1 in Figure 8) was attributed previously to the reduction of the underlying U^IV_1−2xU^V_2x O_2+x layer [21,22], whereas the second peak located at −0.8 V (peak 2 in Figure 8) was assigned to the reduction of UO₃·yH₂O deposit [15].

The current peaks can be considered as a measure of the thickness of the oxidized surface layer. Hence, the thickness of the U^IV_1−2xU^V_2x O_2+x/U^VIO₃·yH₂O layer was lower as the Gd₂O₃ doping level increases. These results are in agreement with the current densities registered in the potentiostatic oxidation experiments (Figure 5).

2.5. Corrosion Potential

Figure 9 shows the corrosion potential (E_CORR) registered on each electrode. The horizontal line at −0.4 V represents the threshold for the onset of fuel corrosion. For all electrodes, the values of E_CORR increased quickly from the cathodic cleaning potential (−1.6 V) until achieving a steady state value. The values of E_CORR recorded after 24 h were −0.10 V, −0.14 V and −0.27 V for undoped UO₂ electrode, 5 wt.% Gd₂O₃ and 10 wt.% Gd₂O₃ doped UO₂ electrodes, respectively, which are substantially more negative as the doping level increases, indicating an influence of the electrode composition on the E_CORR.

Figure 10 shows the XPS analysis performed after E_CORR measurements and

Table 2. Relative fractions of U^IV, U^V and U^VI on the surface of the different samples from XPS results after corrosion potential experiments.

Sample	UIV(%)	UV(%)	UVI(%)
UO2	38 ± 5	48 ± 5	14 ± 5
UO2—5 wt.% Gd2O3	49 ± 5	47 ± 5	4 ± 5
UO2—10 wt.% Gd2O3	87 ± 5	13 ± 5	0 ± 5

displays the fraction of the individual oxidation states of uranium (expressed as percentages) on the electrode surfaces. U^IV is the dominant oxidation state present in the 10 wt.% Gd₂O₃ doped UO₂ electrode surface (around 87%), whereas both undoped and 5 wt.% Gd₂O₃ doped electrodes surfaces are predominantly composed of the oxidation states U^IV and U^V, being the U^IV content slightly higher in the doped electrode than in the undoped one. A low amount of U^VI was detected in UO₂ electrode (14%) while in doped electrodes the amount was almost negligible (below 4%). As mentioned before, silicate and calcium could be adsorbed on the electrode surface, which could partially inhibit the further oxidation of the thin surface layer [19] at such conditions.

Shoesmith et al. reported a reaction sequence on unirradiated UO₂ electrodes in basic solutions in terms of E_CORR [27]. For E_CORR < −0.1 V, the oxidation of UO₂ surface occurs, as a thin surface layer U^IV/U^V is being formed. Therefore, it is expected that the oxidized states U^IV and U^V dominates the surface composition. The surface becomes more oxidized as E_CORR increases. For E_CORR ≥ −0.1 V, the oxidative dissolution of the oxidized layer (as UO₂²⁺) occurs, hence the presence of U^VI is expected as no complexing agents like bicarbonate are present in the solution. For E_CORR > 0 V, the formation of secondary phases (in the form of UO₃·H₂O) is expected to take place on the electrode surface.

According to our XPS measurements, the composition of the surface obtained after the corrosion experiments are consistent with E_CORR values registered at steady state. For undoped UO₂ electrode, the presence of U^VI can be explained due to oxidative dissolution explained for E_CORR ≥ −0.1 V. For doped electrodes, only the oxidation of UO₂ surface occurs, as no U^VI was significantly measured (E_CORR < −0.1 V). However, in the 10 wt.% Gd₂O₃ electrodes, only a slight oxidation of the UO₂ surface (U^V composition of 13%) was observed, which is in agreement with the low E_CORR value obtained (−0.27 V, close to the oxidation threshold). On the contrary, the 5 wt.% Gd₂O₃ doped UO₂ electrode registered a higher E_CORR (−0.14 V) and a higher degree of oxidation (U^V composition of 47%).

These results would confirm that the formation of the thin surface layer U^IV/U^V was influenced by Gd^III-doping in the presence of silicate and calcium. Such retardation of the oxidation process can be attributed to the ability of Gd to delay the incorporation of O^2- into interstitial sites within the fluorite lattice of the UO₂ surface [7,15]. This effect was particularly marked at high doping levels (10 wt.% Gd₂O₃). However, despite the similar E_CORR values achieved for undoped UO₂ and UO₂ doped with 5 wt.% Gd₂O₃, a more rapid oxidation on the undoped UO₂ electrode was observed (the rate of rise of E_CORR was higher for the undoped electrode), evidencing a clear decrease in reactivity due to Gd₂O₃.

3. Materials and Methods

3.1. Electrode Preparation

Undoped UO₂ and UO₂ doped with 5 wt.% and 10 wt.% Gd₂O₃ pellets were fabricated by mixing fine-grained powders of UO₂ (supplied by ENUSA, Madrid, Spain) and Gd₂O₃ powders (99.9%, supplied by Sigma Aldrich, St. Louis, MO, USA). The powders were compacted into pellets by means of a hydraulic press and sintered in a horizontal tube furnace (ST196030 HG from Hobersal, Barcelona, Spain) at 1740 °C for 9 h in a reducing atmosphere (5% hydrogen in 95% argon), based on the methodology explained by Baena et al. [28]. This methodology was confirmed to be adequate for obtaining successfully sintered Gd₂O₃-doped UO₂ pellets [12]. The pellets, with a thickness of 3 mm and a diameter of 12 mm, were characterised by scanning electron microscopy (SEM) and X-ray diffraction (XRD). Then, to study the electrochemical behaviour of the pellets, electrodes were prepared from the pellets by following the procedure described in the literature [29]. First, in order to improve the electrical conductivity, one face of the pellet was electroplated with copper. For that purpose, the pellet was fitted at the end of a rubber tube, with the face to be plated exposed to a solution of 0.1 mol·L⁻¹ CuSO₄ and a Cu metal wire immersed as a cathode. Mercury (Thermo Scientific, Waltham, MA, USA) was poured into the rubber tube to make an electrical contact between the pellet and a stainless-steel wire, acting as an anode. A 10 mA current was applied for 5 min by using a laboratory power supply Manson EP-613 (Kwai Chung, Hong Kong). The Cu-plated side of the pellet was connected to a stainless-steel threaded post by using a conductive silver epoxy (8331-14G). The pellet was embedded with a non-conductive epoxy EE 9466 resin so that only one side of the electrode is in contact with the solution. The electrode was left to cure overnight in a vacuum desiccator. The threaded stainless-steel post was screwed onto the shaft of a rotating disk electrode (Metrohm, Herisau, Switzerland). Both the electrode and stainless-steel shaft were tightly wrapped with poly(tetrafluoroethylene) (PTFE) to avoid entering in contact with the solution. Prior to any experiment, the electrode was abraded using wet SiC paper (1200 grit) and rinsed with distilled deionized water to remove any oxidized surface phase.

3.2. Electrochemical Cell and Equipment

Electrochemical experiments were conducted in a single compartment corrosion cell Autolab 1 L model from Metrohm (Herisau, Switzerland). A standard three-electrode set-up was used with some electrodes of UO₂ (doped with 0, 5 and 10 wt.% Gd₂O₃) as working electrode, the counter electrode was a platinum sheet (surface area ∼1 cm²) and the reference electrode was an Ag/AgCl (3 M KCl) electrode. All potential values are recorded versus the Ag/AgCl standard reference electrode (+0.208 V, 25 °C vs. standard hydrogen electrode (SHE)). Experiments were conducted at room temperature. All electrochemical experiments were performed using Multi Autolab Potentiostat-Galvanostat M204 from Metrohm, (Herisau, Switzerland). NOVA software 2.1 (Metrohm, Herisau, Switzerland) was used to record and to analyse the data.

3.3. Electrochemical Techniques

Four general types of electrochemical techniques were carried out: cyclic voltammetry (CV), potentiostatic oxidation, cathodic stripping voltammetry (CSV) and corrosion potential (E_CORR). Prior to any experiment, the electrode was cathodically cleaned for 120 s at −1.6 V to remove U^VI oxides present on the surface. CV is a potential tool to investigate the potential at which UO₂ undergoes oxidation or reduction in a solution. In that regard, the potential (V) was scanned from −1.6 V to 0.6 V and back at a rate of 10 mV·s⁻¹ while recording the current. The electrode rotation rate was held constant at 1000 rpm. In potentiostatic oxidation experiments, a constant anodic potential of 0.4 V was applied to the electrode for an hour and current-time profiles were recorded. This technique is capable of identifying variations in oxidation rates and levels of the electrode at a specific oxidizing potential due to Gd doping. After the anodic oxidation, the surface of the electrode was analysed by XPS. Immediately after a second set of identical potentiostatic oxidation experiments, cathodic stripping voltammetry was conducted in order to reduce the oxidized surface layer formed on the electrode. The cathodic charge measured can serve as an indicator of the layer thickness. This was performed by scanning the potential from the potentiostatic oxidation potential (0.4 V) to −1.6 V at a scan rate of 10 mV·s⁻¹, while the current was registered. Finally, in the corrosion potential experiments, the working electrode was immersed in the solution for 24 h, while the potential that exists across the electrodes and electrolyte was measured in the absence of any current flow. This measurement serves as an indicator of the tendency of an electrode to corrode in a specific environment and XPS analysis was performed after the E_CORR measurement.

3.4. Solution

The solution used for all experiments was 0.1 mol·dm⁻³ NaCl (99.0%, Sigma Aldrich, St. Louis, MO, USA) as a supporting electrolyte, containing 10⁻² mol·dm⁻³ SiO₃²⁻ (Na₂SiO₃·5H₂O (99.0%, Sigma Aldrich, St. Louis, MO, USA)) and 10⁻³ mol·dm⁻³ Ca²⁺ (CaCl₂ (99.0%, Sigma Aldrich, St. Louis, MO, USA)). In addition, for testing the performance of the electrodes in cyclic voltammetry experiments, a solution that contained 0.1 mol·dm⁻³ NaCl (99.0%, Sigma Aldrich, St. Louis, MO, USA) and 5·10⁻² mol·dm⁻³ of NaHCO₃ (99.0%, Sigma Aldrich, St. Louis, MO, USA) was prepared. All solutions were prepared using ultrapure water obtained from a Milli-Q water purification system (Merck, Darmstadt, Germany). The pH of the solutions was adjusted to 10 using HCl or NaOH and measured with an Orion model 720A pH meter (Waltham, MA, USA). The pH was measured periodically and did not vary more than 1% throughout the experiments. The solutions were bubbled with N₂ (99.99%, Nippon Gases, Tokyo, Japan) before any experiment for 1 h to purge and during the experiments.

3.5. XPS Analysis

XPS analyses were performed to observe changes in the oxidation of the electrode surface after potentiostatic oxidation and corrosion potential experiments. A SPECS system equipment (Berlin, Germany) with an Al anode XR50 source operating at 150 W and a Phoibos MCD-9 detector was used. The sensitivity of the equipment was ±0.1 eV of binding energy (BE). The spectra were recorded at a pressure below 10⁻⁸ mbar. The spectra were evaluated using CasaXPS software (Casa Software Ltd., Teignmouth, UK) version 2.3.25. C 1s peak was set at 285.0 eV to correct surface charging. The spectra were fitted using a 30% of Lorentzian curve and 70% of Gaussian curve with a Shirley background correction. The U4f_7/2 and U4f_5/2 regions were used to determine the total amount of each uranium oxidation state (U^IV, U^V and U^VI) by the deconvolution of the peaks [30,31,32,33,34]. The position of the satellite peaks was used to validate the curve-fitting procedure.

Before transferring to the XPS spectrometer for XPS measurements, electrodes were removed from the electrochemical cell, gently rinsed with MilliQ water to remove electrolyte residue and dried in a desiccator.

3.6. SEM Measurements

A Phenom XL scanning electron microscope coupled to an energy dispersive X-ray detector (EDX) from PhenomWorld (Waltham, MA, USA) was used to analyse the microstructure of the samples.

3.7. X-ray Diffraction (XRD)

A Bruker D8 diffractometer (Karlsruhe, Germany) with an X-ray source of Cu Kα radiation (λ = 1.54056 Å) at 45 kV and 35 mA was used to analyse the crystal structure of the samples.

4. Conclusions

The effect of Gd (5 wt.% and 10 wt.% Gd₂O₃) on the electrochemical behaviour of UO₂ was investigated in a slightly alkaline solution in the presence of silicate and calcium. Cyclic voltammetry experiments showed that the anodic oxidation/dissolution mechanism on Gd₂O₃-UO₂ was similar to that on undoped UO₂ electrode. However, the corrosion process of doped UO₂ decreases, as lower current densities were registered. Both potentiostatic and E_CORR experiments showed that such retardation effect was small when doping with 5 wt.% Gd₂O₃, however, it was strongly enhanced when Gd₂O₃ content is around 10 wt.% Gd₂O₃. This behaviour is commonly attributed to the reduction in the availability of oxygen vacancies to accommodate O²⁻ ions during oxidation in the UO₂ lattice, as Gd^III-O vacancy clusters might be formed.

In addition, the presence of silicate and calcium in solution was found to have influence on the oxidation of UO₂, leading to a suppression of anodic dissolution probably due to the adsorption of these ions on the electrode surface. This is in contrast to solutions that contain bicarbonate ions, which are known to form soluble complexes with UO₂²⁺ and to accelerate UO₂ corrosion. Additional experiments with solution containing Ca, Si and carbonate would be required to investigate which effect would predominate.

In conclusion, both Gd₂O₃ doping and the presence of silicate and calcium in solution are factors that decrease the reactivity of UO₂, which is beneficial for the assessment of repository safety, as an important fraction of the radionuclide release would be inhibited.