Quantitative Analysis of Oxygen Gas Exhausted from Anode through In Situ Measurement during Electrolytic Reduction
Quantitative analysis by in situ measurement of oxygen gas evolved from an anode was employed to monitor the progress of electrolytic reduction of simulated oxide fuel in a molten Li2O–LiCl salt. The electrolytic reduction of 0.6 kg of simulated oxide fuel was performed in 5 kg of 1.5 wt.% Li2O–LiCl molten salt at 650°C. Porous cylindrical pellets of simulated oxide fuel were used as the cathode by loading a stainless steel wire mesh cathode basket. A platinum plate was employed as the anode. The oxygen gas evolved from the anode was exhausted to the instrumentation for in situ measurement during electrolytic reduction. The instrumentation consisted of a mass flow controller, pump, wet gas meter, and oxygen gas sensor. The oxygen gas was successfully measured using the instrumentation in real time. The measured volume of the oxygen gas was comparable to the theoretically calculated volume generated by the charge applied to the simulated oxide fuel.
Electrolytic reduction (also called oxide reduction or OR) of solid compounds to solid products in molten salt electrolytes has attracted widespread interest in metallurgy because it is simple, cost-effective, and environmentally friendly [1–3]. This technique has been applied to the reduction of various metal oxides such as TiO2 , SiO2  Ta2O5 , Fe2O3 , SnO2 , Tb4O7 , Nb2O5 [10–12], Cr2O3 [13, 14], and CeO2 . Additionally, it has been employed to reduce spent oxide fuel (mainly consisting of UO2) to its metallic form in pyroprocessing technology for closed nuclear fuel cycles [16–18]. Pyroprocessing is a high-temperature electrochemical fuel processing technology for recycling the spent oxide fuel from light water reactors into metal fuel for fast nuclear reactors. In the OR process for pyroprocessing, molten LiCl containing Li2O is generally used as the electrolyte for the reduction of the spent oxide fuel. The reduction products of the OR process are transferred to an integrated process, electrorefining, as feed materials. U, Pu, and other actinides are recovered by the electrorefining process in a LiCl–KCl eutectic salt electrolyte at 500°C [19–26].
The electrode reactions in OR processes are well known [16–18, 27–32]. The spent oxide fuels are used as cathodes by loading them into a permeable basket. The reactions at the cathode are as follows:where denotes actinides such as uranium and plutonium. Simultaneously, the main reaction at the Pt anode occurs as follows: During electrolysis, the spent oxide fuel is reduced to metal and remains at the cathode. The ions produced at the cathode are then transported through the salt and discharged at the anode to form O2 gas.
In situ monitoring of the OR process enables efficient operation; its principle purpose is to monitor the performance and efficiency of a reduction cell. In particular, the extent to which the oxide fuel is converted to metal at the cathode is key information for determining the efficiency of a reduction cell. However, in situ monitoring of the reduction extent in real time has not been realized to date. Alternatively, the cumulative electrical charge applied to the oxide fuel ( [C]) which is obtained from the current () and time (t [s]) during electrolysis is typically monitored. If the efficiency of the reduction cell is 100%, then is equal to the theoretical amount of electrical charge () required to reduce all of the metal oxide to metal. However, since not all of is used to convert oxide to metal and the efficiency of the reduction cell is less than 100%, the fraction of relative to is ambiguous. Here, we test the feasibility of quantitatively measuring O2 gas as a means of in situ monitoring of the OR process. The advantage of this approach is that it allows monitoring not of the applied quantity but of the produced quantity from OR reactions. For this study, the OR of 0.6 kg of simulated oxide fuel (simfuel) is performed in molten Li2O–LiCl salt. The O2 gas produced at the anode is exhausted and quantitatively measured by means of in situ measurement instrumentation.
2. Experimental Process
The OR apparatus was used in a high-purity argon (Ar) atmosphere glove box (Figure 1(a)) where the concentrations of moisture and O2 were kept at below 1 ppm and 10 ppm, respectively. Figure 2(a) shows a schematic drawing of the OR apparatus used in the present study. It consists of several components including a furnace (① in Figure 2(a)), outer STS crucible (② in Figure 2(a)), inner STS crucible (③ in Figure 2(a)), top flange (④ in Figure 2(a)), and electrodes (⑤, ⑥, and ⑦ in Figure 2(a)). The inner STS crucible (165 mm × 165 mm × 250 mm) was used to load the electrolyte. As the OR electrolyte, 5 kg of 1.5 wt.% of Li2O (99.5% purity, Alfa Aesar)–LiCl (99% purity, Alfa Aesar) molten salt was used at 650°C. LiCl was thermally dehydrated before use to remove moisture. The outer STS crucible aimed to protect the furnace in case of failure of the inner STS crucible. The top flange above the crucibles provided ports to suspend the electrodes and also prevented heat loss of the electrolyte.
The electrodes were electrically insulated from the top flange using a ceramic material at the contact area (⑧ in Figure 2(a)). The cathode (⑥ in Figure 2) was used by assembling the Ni conductor container and the rectangular basket (75 mm × 25 mm × 204 mm) made of three-ply layered (20-325-100, Nichidai)-STS wire meshes. Porous simfuel pellets with a tap density of 3.59 g/cm3 (density of single pellet of 6.89 g/cm3, bulk density 62.86%) and a cylindrical shape (ϕ6.4 × 6.2 mm) were used by loading the STS wire mesh cathode basket. Its components and composition are listed in Table 1; it was composed of UO2, rare earth oxides, noble metal oxides, and salt-soluble fission products. The simfuel was prepared by mixing of UO2 and surrogate powders, pelletizing, and sintering. The preparation conditions and methods for simfuel have been described elsewhere . Li–Pb alloy (32 mol% Li) in MgO served as the reference electrode (⑦ in Figure 2(a)). The anode module (⑤ in Figure 2(a)) consisted of a Pt anode and a shroud. A Pt plate (40 mm × 3 mm × 200 mm) was used as the anode material. The anode shroud surrounds the Pt anode where O2 gas is generated; it forms chimneys that contain the O2 gas and form a pathway for its removal. The anode shroud used in this study consisted of a lower porous shroud for the salt phase and an upper nonporous shroud for the gas phase. Detailed descriptions of the anode and anode shroud have been reported in our previous study . As shown in Figure 2(b), the evolved gas combination of O2 and Ar was continuously exhausted using a pump from the upper end of the shroud to the gas outlet; the gas was piped to in situ instrumentation installed outside the Ar glove box. The setup of the O2 gas monitoring instrumentation is presented in Figures 1(b) and 1(c). The instrumentation consisted of a mass flow controller (MFC), pump, wet gas meter, and sensor for O2 gas. The MFC (Line Tech, m3030v, Korea) was used to control the flow rate of the gas using the connected pump; it was given a set point of 1 L/min. The removal rate of O2 gas was set to be faster than the generation rate. As shown in Figure 2(b), Ar gas was allowed to flow into the top flange in order to constantly maintain the pressure inside of the OR apparatus. The wet gas meter (Shinakawa, W-NK 1B type, Japan) was employed to quantitatively measure the total volume () of the evolved gas (a combination of O2 and Ar gases) in real time. The volume of O2 () was determined by multiplying by the O2 concentration fraction () measured using an oxygen sensor (Advanced Micro Instruments, M65, US). was monitored and recorded in real time using a computer. The O2 gas concentration in the Ar-filled glove box was carefully observed in order to determine whether O2 gas leaked out from the OR apparatus or the connected piping. Electrolysis was performed by applying a constant voltage between the cathode and Pt anode using an external power supply (Agilent Model 6671A). The potential between the cathode and the Li–Pb reference electrode was monitored using a multimeter (Agilent, 34405A). After the OR process, the variation of Li2O concentration in the molten salt was checked using an autotitrator (G20, Mettler Toledo). A filter composed of a cylindrical polypropylene microfilter with a pore size of 1 μm was used to protect the instrumentation from salt aerosols.
3. Results and Discussion
The constant-voltage electrolysis of 0.6 kg of simfuel was performed in 1.5 wt.% Li2O–LiCl. A set of the obtained electrolysis data is presented in Figure 3; it includes the curves of the (a) cell voltage, (b) current, and (c) cathode potential. The constant voltage measured between the tops of the cathode and the anode was approximately 3.15 V owing to the IR drop, even though 3.35 V was applied by using the power supply (Figure 3(a)). The constant-voltage interruption at appropriate intervals was conducted to prevent excessive formation of Li metal at the cathode. The presence of Li metal at the cathode could be observed through the potential between the cathode and Li–Pb reference electrode at an open circuit potential (i.e., the Li/Li+ potential of about −0.6 V). While the constant voltage was applied, the cathode potential was lowered to about −0.78 V (Figure 3(c)) because, as a cyclic voltammetry test in our previous study showed, UO2 is reduced at potentials more negative than −0.6 V [30, 31]. A response current of about 40 A was passed by applying the constant voltage (Figure 3(b)). The averaged current density was 0.45 A/cm2 for an effective anode surface area of 89.2 cm2 immersed in the Li2O–LiCl salt. relative to 40% of was the applied charge to the simfuel during the 6.3 h electrolysis, which was not sufficient to completely reduce the simfuel. This incomplete electrolysis (low compared to ) was inevitable to monitor the O2 gas using the instrumentation without interruption during the daytime.
The evolved gas generated from the OR anode was continuously exhausted and monitored during the electrolysis. Figure 4 shows the cumulative measured volume of O2 gas () compared with the theoretically calculated volume of O2 gas () generated by the cumulative applied charge, . was obtained from the current and time shown in Figure 3(b). The change in showed a step-like pattern; it repeatedly increased and then maintained a constant value because the current flowed during the voltage-on period and did not flow during the voltage-off period. was calculated as follows: where is the electrical charge applied to the simfuel during electrolysis; is the number of electrons involved in the cathode reaction (1) (known to be four); is Faraday constant, 96,485 C/mol e-; and 24.5 is the molar volume of ideal gas (L/mol) at 25°C.
As expected, increased in the same manner as because it was obtained using in (5). The change in also showed a similar pattern to that of , which is evidence of successful production of O2 gas by the anode reaction (4). At the beginning of the electrolysis, did not show a significant increase in spite of the increase in ; it increased after about 10 min. This 10 min interval between and was repeatedly observed during the electrolysis. This interval indicates the time needed for the O2 gas generated in the molten Li2O–LiCl salt of the OR apparatus installed inside the Ar glove box to pass through the piping and to be measured in the instrument installed outside the glove box. It is notable that the difference between and gradually increased over the electrolysis period. At the end point of the electrolysis, of 16 L was approximately 73% of of 22 L that could theoretically be generated by the final value of the cumulative (3.4 × 105 C). This result is reasonable given the fact that is not all used to convert the oxide fuel to metal owing to limited current efficiency. In our previous study , we conducted the OR process using simfuel under experimental conditions similar to those in the present study. We obtained the reduction product containing U with the metal weight portion of 99% when the OR process with a current efficiency of 67% was conducted. This reliability was confirmed from the observation of the reduction product obtained from the OR run of this study. Some reduction products were randomly sampled from the top of the cathode basket; a cross-section of one of them is shown in the inset image of Figure 5(a). Its surface exhibits a metallic gray appearance unlike the original brown simfuel, which is also evidence of successful reduction during electrolysis. However, the unreduced region in the center of the cross-section remained because (40% of ) was not enough for complete reduction. This inhomogeneity of the reduction over the region occurs because reduction of simfuel begins at the surface and progresses toward the center . The concentration of Li2O (1.5 wt.%) in LiCl measured through titration did not show a change before and after electrolysis, which means that the O2 gas measured using the instrumentation () was generated from the metal oxide in the simfuel.
The electrolysis was also evaluated by visual inspection of the reduced simfuel pellets. After electrolysis, the electrodes were removed from the molten Li2O–LiCl at 650°C and cooled to the ambient temperature. The used anode shroud (Figure 5(b)), Pt anode (Figure 5(c)), inner crucible (Figure 5(d)), and top flange (Figures 5(e) and 5(f)) were inspected. No severe corrosion or damage was observed, which indicates that the O2 gas was successfully discharged along the intended path without a leak in the OR apparatus and measured using the instrumentation.
The feasibility of quantitative measurement of the O2 gas generated from a Pt anode was tested to monitor the progress of the OR process. In situ measurement of O2 gas was performed using monitoring instrumentation during the OR of 0.6 kg of simfuel in molten Li2O–LiCl salt. The measured cumulative volume of O2 gas () increased with the electrolysis time in a pattern similar to that of the cumulative applied charge, which implies successful monitoring of OR progress. The measured volume of the O2 gas using the applied electrical charge () amounted to around 73% of the theoretical calculated volume () at the end point of the electrolysis. This OR progress was also confirmed by visual inspection of the reduced simfuel. This approach is advantageous because it allows the monitoring of not the applied quantity but the produced quantity from OR reactions. However, the leakage of O2 gas in the instrumentation should be carefully avoided for its successful monitoring.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
This study was supported by the National Research Foundation of Korea (NRF) of the Korean Government (MISP) [Grant no. 2017M2A8A5015077].
L. I. Redey and K. Gourishankar, “Direct electrochemical reduction of metal-oxides,” US Patent 6540902B1.View at: Google Scholar
S. D. Herrmann, S. X. Li, M. F. Simpson, and S. Phongikaroon, “Electrolytic reduction of spent nuclear oxide fuel as part of an integral process to separate and recover actinides from fission products,” Separation Science and Technology, vol. 41, no. 10, pp. 1965–1983, 2006.View at: Publisher Site | Google Scholar
R. W. Benedict and H. F. McFarlane, “EBR-II spent fuel treatment demonstration project status,” Radwaste Magazine, vol. 5, p. 23, 1998.View at: Google Scholar
S.-C. Jeon, K.-Y. Lee, J.-W. Lee, Y.-Z. Cho, and D.-H. Ahn, “Scaling up and tailoring the fabrication of UO2 porous pellet with a simulated spent fuel composition,” in Proceedings of the International Pyroprocessing Research Conference, Jeju Island, Korea, 2016.View at: Google Scholar