1.Introduction

Pyroprocessing is currently under development to recycle and manage the spent oxide fuels generated from nuclear reactors [1-3]. Electrolytic reduction is the first electrochemical process in pyroprocessing, in which oxide fuels are converted to the metallic states. Then the reusable nuclear species (e.g., U, Pu) are recovered during subsequent electrorefining and electrowinning processes for use as metal fuels in next-generation fast reactors [1-3].

Li₂O-containing LiCl molten salt at 650℃ is commonly used as an electrolyte for electrolytic reduction in pyroprocessing [4-10]. During the electrochemical reaction, metal oxides (e.g., UO₂) are reduced to metals (e.g., U) at the cathode, while O₂ gas is formed at the anode [4-5]. The anode is thus exposed to a highly oxidizing environment because of the high reaction temperature. Although Pt is the most widely used anode material because of its resistance against oxidation [4-7], a Pt anode would be damaged during the electrochemical reaction owing to the formation of Li₂PtO₃ and anodic dissolution [5]. In our previous papers, we have suggested various materials as alternatives to expensive Pt anodes [8-10]. Carbon-based compounds and liquid Sb anodes have shown low current efficiency, however, because of side reaction cycles during the reaction [8-9]. While B-doped diamond was able to reduce UO₂ in a KCl–LiCl–Li₂O electrolyte at a low temperature of 550℃, it had poor current density [10].

In this study, we studied the feasibility of a conductive nitride, TiN, as the anode material. TiN thin films have been widely investigated as an electrode material in semiconductors because of their thermal stability in an oxidizing atmosphere [11-12]. During the electrochemical reaction, TiN would be sacrificially oxidized to form oxides and N₂ gas (2TiN + 2O₂(g) = 2TiO₂ + N₂(g), ΔG = –1052.15 kJ at 650℃), which is not corrosive. Alternatively, it may remain stable during the reaction because the oxidation kinetics is significantly affected by the O₂ partial pressure. Indeed, the oxidation behavior of TiN varies under different oxidation conditions [13-14]. Herein, we investigated the electrochemical properties of the TiN anode in a 650℃ LiCl–Li₂O electrolyte that was used to reduce UO₂ to metallic U.

2.Experimental

A disk-shaped bulk TiN of 50.8 mm in diameter and 6.35 mm in thickness (Alfa Aesar, USA) was cut into adequate dimension (1/4 position from edge) and then connected to stainless steel lead for the test (Fig. 1(a)). An electrolytic cell was installed inside an Ar-filled glove box. Detailed descriptions of the experiments are provided in the literature [4-7]. Prior to the electrochemical study, the TiN anode was immersed in a 700 g of LiCl–Li₂O (1 wt% Li₂O) electrolyte at 650℃ to examine its stability in the molten salt. The Li₂O electrolysis test was done with a stainless steel cathode (rod-type) and the TiN anode. A stainless steel basket containing 10 g of UO₂ was used as the cathode for the electrolytic reduction test. A constant voltage was applied using a power supply (E3633A, Agilent, USA) during the reaction and, at the same time, the potential difference between the cathode and the Li–Pb reference electrode was monitored using a digital multimeter (34401A, Agilent, USA).

The crystallinity was determined by X-ray diffraction (XRD; D8 Advance, Bruker, Germany). Thermogravim_tery (TG; TGA/DSC1 Mettler Toledo, Switzerland) was used to evaluate the conversion rate of UO₂ to metallic U, based on the re-oxidation of U (or un-reduced UO₂) to U₃O₈ in O₂ atmosphere (3U + 4O₂(g) = U₃O₈ or 3UO₂ + O₂(g) = U₃O₈). Prior to the TG measurement, the reaction product was rinsed with distilled water to remove residual LiCl-Li₂O. Elemental analysis was carried out by energy-dispersive X-ray spectroscopy (EDS; X-MAX, Horiba, Japan) coupled with scanning electron microscopy (SU-8010, Hitachi, Japan).

3.Results and discussion

Before the electrolytic reduction test, a preliminary test on the stability inside the molten salt and capability of Li₂O electrolysis was carried out, as shown in Fig. 1. The TiN anode sustained its shape after immersion in the molten salt electrolyte (Fig. 1(a) and (b)), demonstrating the stability of TiN. No breakdown induced by thermal shock, which is frequently observed in ceramic compounds, was seen in the TiN anode. Fig. 1(c) shows the Li₂O electrolysis behavior of the TiN. The decomposition of Li₂O is an important process in the electrolytic reduction owing to the catalytic behavior of Li₂O [4-7]. Once Li₂O dissociates to form metallic Li on the cathode and O₂ gas on the anode (2Li₂O = 4Li + O₂(g)), the metallic Li readily reacts with UO₂ to form metallic U and new Li₂O (UO₂ + 4Li = U + 2Li₂O) [4-7]. The stainless steel rod without UO₂ was used as the cathode here. The cell voltage between the anode and the cathode was controlled to vary from 0.5 to 3.25 V to determine adequate reaction voltage range. When the cell voltage was above 3.0 V, the cell current increased drastically, which means that the electrochemical reaction was triggered between 2.5 and 3.0 V. High reaction voltage compared the decomposition voltage of Li₂O (~2.46 V) might be due to the surface overpotential, induced by O₂ bubble, electric double layer, or some other reasons. The cathode potential (i.e., potential difference between the cathode and the reference electrode) at this range was approximately –0.73 and –0.56 V vs. Li–Pb in the closed-circuit and open-circuit configuration, respectively. The open-circuit potential of the cathode is comparable to that of the Li⁺/Li⁰ redox couple (approximately –0.54 V vs. Li/Pb) [4], demonstrating the formation of metallic Li on the cathode surface. The more negative value of the closed-circuit potential was due to the polarization applied to the cathode. The metallic Li deposited on the cathode rod surface can be clearly seen in the inset of Fig. 1(c).

Fig. 2 shows the electrolytic reduction behavior of the TiN anode with the UO₂-containing stainless steel cathode basket. The cell voltage was fixed at 3.25 V to induce the electrolysis of Li₂O, as shown in Fig. 2(a). A value equivalent to 150% of the theoretical charge to convert all UO₂ to U was supplied to the electrolytic cell. The open-circuit potential is comparable to that shown in Fig. 1(c), indicating the formation of Li. The lower cell current compared to that observed in the Li₂O electrolysis test was due to the reduced cathode area. After the reaction, the color of the sample turned from dark brown of the as-prepared UO₂ to shiny silver of the reduced UO₂ (Fig. 2(b)), suggesting that the electrochemical conversion of UO₂ to the metallic state took place. We carried out XRD analysis to identify the reaction product, and the result is shown in Fig. 2(c). The diffraction peaks are well matched to the metallic U phase with small UO₂ peaks. The conversion rate was estimated to be 87.11–100.58% by TG analysis at different locations of the reduced UO₂ sample. These results obviously confirm that the TiN anode was capable of electrochemically reducing metal oxides to their metallic forms.

After the reaction, no oxidation of the TiN anode was observed in the XRD pattern in Fig. 3(a). On the other hand, a number of voids were formed in the immersed area of TiN, while no noticeable change was seen in the non-immersed area (Fig. 3(b) and (c)), which emphasize the electrochemical instability of the TiN anode after prolonged reaction duration. A shiny layer also can be seen on the surface of the TiN anode. This is due to the corrosion of the stainless steel lead, originated from the surface climbing of the molten salt through the electrode surface. The degradation mechanism of TiN remains unclear at this moment, but it is speculated that the electrolyte-soluble phase, somehow, was formed during the reaction. Indeed, EDS analysis of the electrolyte byproduct after the reaction shows the dissolution of Ti into the electrolyte (Fig. 3(d)). Despite the void formation, a dominant reaction pathway is believed to be the O₂ evolution reaction. Fig. 3(e) shows the bottom plate of the electrolytic reducer flange made with stainless steel. The stainless steel flange became significantly rusty, implying the formation of a large amount of O₂ during the reaction.

4.Conclusion

The suitability of the conductive nitride TiN as anode material for the electrolytic reduction in pyroprocessing was investigated. The TiN anode was shown to electro chemically reduce UO₂ to metallic U in a molten salt electrolyte. While no noticeable oxidation of the TiN anode was observed, the dissolution of TiN gradually occurred during the reaction. Hence, it is apparent that TiN has a limited lifetime as the anode material in the electrolytic reduction and its use should be limited in small-scale or short-term experiments.