1.Introduction

Pyroprocessing is under intensive research at the Korea Atomic Energy Research Institute (KAERI) as a used nuclear fuel (UNF) management method [1-2]. A recent agreement between the United States and the Republic of Korea to perform head-end and electro-reduction processes at KAERI using real UNFs opened up an opportunity for more realistic research into pyroprocessing. In addition to proliferation resistance, pyroprocessing also has a merit of a reduced amount of high-level waste enabling an efficient repository use. However, pyroprocessing still generates some radioactive wastes including off-gas (volatile and semi-volatile nuclides), metal (fuel assembly structure and cladding hulls), and salt wastes (LiCl and LiCl-KCl salts employed in electro-reduction and electro-winning, respectively). Therefore, waste streams need to be clarified to complete the pyroprocessing at a commercial scale.

In the waste streams of pyroprocessing, iodine is one of the key nuclides that need to be managed carefully to achieve environmental acceptance and the reliability of the process equipment. UNFs are cooled down for years before being charged for the pyroprocessing, and ¹²⁹I will remain as a major radioactivity source of iodine owing to its long half-life of 15.7 million years. This extremely long halflife of ¹²⁹I makes it clear why the management of iodine is important. In addition, it should be noted that the high reactivity of iodine with metals will cause a serious damage to the reactors and electrode materials of the subsequent electrochemical processes. Thus, in the pyroprocessing of KAERI, it is planned that most of iodine is removed from UNF through heat treatments in the head-end process so that iodine cannot hinder the subsequent electrochemical reactions.

Techniques to capture iodine have been intensively researched based on wet reprocessing techniques [3]. The most successful method was liquid solvent scrubbing, which employs reactive chemical solutions to hold iodine within the liquid solvent. This method requires further steps such as chemical conversion of iodine and stabilization before permanent disposal. As an alternative technique, a dry scrubbing process was also investigated especially for silver based solid adsorbents. Normally, silver is ion-exchanged on large surface area zeolites resulting in about 15 wt% silver contents [3]. However, we need to discuss some of the disadvantages of a silver-zeolite adsorbent from a waste management point of view. Firstly, employing a large surface area zeolite will increase the chance to capture iodine physically unless a high temperature is applied during the capture. Physically adsorbed iodine is troublesome for high-temperature waste form fabrication processes. Secondly, owing to limitations in increasing the silver content, the silver-zeolite adsorbent will generate a bulky waste form. Surely this disadvantage can be covered by leaching AgI or iodine, but it requires an additional process. Thirdly, the high price of silver-zeolite adsorbent should be mentioned. Maybe recycling can be performed by treating AgI/ zeolite to selectively recover the iodine, but an iodine recovery and chemical conversion process might be additionally required. However, silver-zeolite adsorbent is still the hottest material as a dry scrubbing media thanks to its ability to capture methyl iodide (CH₃I) and form stable metal iodide (AgI) in the presence of oxygen [3-5]. Here, a question should be asked regarding whether we need costly silver to capture iodine generated from the pyroprocessing. As pyroprocessing does not use organic solvents, it is unlikely that methyl iodide can be produced. In addition, the removal of iodine will be done through a high-temperature (above 1000℃) heat treatment process during the head-end process, which means that methyl iodide will be decomposed before they meet the capturing medium [6]. The stability of AgI in the presence of oxygen is still beneficial although the high-temperature heat treatment might be done under Ar or H₂/Ar atmosphere, because a small amount of oxygen will be liberated during the conversion of U₃O₈ into UO_2+x [2]. However, a low concentration of oxygen and inert/reducing atmosphere definitely dilutes the merit of silver.

In the present study, a copper mesh was proposed as an iodine capturing adsorbent for the pyroprocessing for the following reasons. First, the mesh structure might not produce physically adsorbed iodine because it does not contain micro/nano-pores. In addition, commercial copper meshes can be purchased at a low price and will produce only copper iodide and copper as its source for waste form fabrication. One disadvantage of employing copper is relatively high solubility in water, which is 4.2 × 10^-4 g/L at 25℃ [7] whereas that of AgI is 3 × 10^-6 g/L at 20℃ [8]. However, this issue can be managed by choosing an appropriate waste form. The possibility of metallic copper was previously demonstrated by Tachikawa and co-workers [9] in 1975 using neutron activated iodine and copper granule as a copper column. In their work, the regeneration of copper was also demonstrated using hydrogen to produce HI and metallic copper [9]. In the present study, we focused on identifying whether copper meshes have the reasonable potential to be explored for further research. As a starting point of copper mesh research, the effect of the reaction temperature was determined and a characterization of the reaction products was performed.

2.Experimental

A schematic diagram of the experimental set-up is shown in Fig. 1. A vertically positioned quartz tube with a quartz frit in the middle was employed as a reactor. A copper mesh (50 mesh, 0.26 g) was put on the quartz frit and heated to a reaction temperature using a tubular heating mantle. Gaseous iodine was fed from the bottom of the reactor by heating a two-neck flask holding 2 g of iodine on the bottom with an argon flow of 100 mL/min. The temperature of the iodine part was set to 130℃, which is high enough to evaporate iodine within a 1.5 h reaction time. Gaseous iodine was mixed with an argon gas feed and moves up to the reaction zone so that it can react with a copper mesh. Un-reacted iodine and argon gas moved to the top of the quartz reactor and were then introduced to a gas treatment system. A crystal structural analysis of the reaction products was conducted using an X-ray diffraction (XRD) technique. A morphology analysis of the reaction products was conducted using the Scanning Electron Microscopy (SEM) technique.

3.Results and Discussion

3.1.Theoretical calculations

Before getting started with the experiments, theoretical calculations were performed to estimate the appropriate reaction temperature for the reaction between copper and gaseous iodine. The ‘Reaction equations’ module of the HSC chemistry code [10] was employed in this work to calculate the Gibbs free energy of the reaction between copper and iodine. Table 1 lists the calculation results achieved using the HSC chemistry code [10]. As shown in the table, both mono- and bi-atomic iodine gases were considered for the calculation. In the temperature range considered for the calculation, it is clear that the formation of CuI will proceed spontaneously. In addition, it needs to be mentioned that the I(g) form is preferred for the CuI formation reaction rather than the I₂(g) form. These results suggest that a reaction might occur easily at a relatively low temperature. Interestingly, it was identified by further calculation that the change in Gibbs free energy for Cu and Ag are close. At 500℃, the reaction between silver and I(g) to produce AgI exhibited a ΔG value of -22.27 kcal, which is very close to -22.11 kcal for CuI formation. When I₂(g) is employed instead of I(g), the result was -13.67 for Ag and was -13.51 kcal for Cu. These results suggest that the use of Cu instead of Ag might not bring a significant difference from a thermodynamic point of view. Here, it needs to be noted that the formation of CuI₂ was not considered in this work, because it is known to be unstable and easily decomposed into CuI and I(g) [11].

3.2.Experimental results

First, the effect of reaction temperature on the iodine capturing efficiency was investigated. The experiments were conducted at various reaction temperatures of 100, 200, 300, and 400℃. The degree of reaction (α) was calculated using the following equation:

where 63.55 and 126.90 are the atomic weight of copper and iodine, respectively. In Eq.(1), m₀ is the initial weight of a copper mesh, and m_t is the weight of the mesh after the reaction. As the reaction temperature increased from 100 to 200, 300, and 400℃, the α value increased significantly from 0.02 to 0.06, 0.19, and 0.23, respectively. These values also mean that the copper meshes captured 0.5, 1.5, 5, and 6wt% of iodine employed in the experiments. Here, it needs to be mentioned that the degree of reaction rapidly increased in the 100 to 300℃ region, while the increase was relatively suppressed in the 300 to 400℃ region. Images of the reaction products with a bare Cu mesh are shown in Fig. 2. Interestingly, the formation of white/dark green powders on the surface of the copper meshes are shown especially for the 300 and 400℃ cases. It is suspected that the powder is copper iodide which includes some impurities, because CuI is known to be white when pure, and tan or brownish when impure. As shown in the figure, the powders were easily seperated from the meshes during sample handling. This behavior might have come from the density reduction by CuI formation, because ideal densities of metallic copper and CuI are 8.96 and 5.67 g/cm³, respectively [10]. In other words, the volume per unit mass of metallic copper will increase by 58% (= (8.96-5.67)/5.67) as it is converted into CuI. This large volume increase can induce significant stress in the Cu-CuI interface leading to a spontaneous peeling-off of the reaction products. The spontaneous peeling-off of CuI is beneficial for achieving high copper utilization values, because a metallic copper surface will be supplied continuously without any external forces until the whole copper is converted into CuI.

To see what happens when the samples go through additional iodine capture, the sample reacted at 400℃ was employed for a repeated reaction under an identical condition. The picture of the copper mesh after the repeated reaction is shown in Fig. 3a, where a formation of a significant amount of powders is clearly observed. After the second reaction, the α value increased from 0.23 to 0.35 meaning that 35wt% of the copper mesh was consumed to capture iodine while 65wt% of copper remained in its metallic form. After the repeated reaction at 400℃, the powders formed on the surface were scrapped using a spatula. The powders were easily detached from the mesh as shown in Fig. 3b, resulting in powder-free copper surface. The mass of the copper mesh after the powder removal was 0.15 g, which is 57.7wt% of its initial value (0.26 g). The amount of the powder form was 0.29 g meaning that the powder is composed of 0.11 g of copper and 0.18 g of iodine. Based on the weight measurement, the atomic ratio between the copper and iodine of the powder is (Cu:I) = (1.00 : 0.82) suggesting that an iodine deficient form of CuI_0.82 was formed through the reaction. This result also means that Eq. (1) which assumed a formation of Cu₁I₁ needs to be revised as follows.

The reaction products were further analyzed using the XRD technique to identify the structural characteristics. Fig. 4 shows the XRD pattern of the detached powders after the 200℃ reaction (Fig. 2c). It is clear from the figure that the XRD peaks of the reaction products match well with those of the reference peaks (cubic CuI with a space group of Fm-3m [12]). However, it needs to be discussed that the measured peaks at a high 2θ region were observed at higher 2θ positions than the reference values. A right-hand shift of the peak positions without an appearance or disappearance of the peaks means a reduced d-spacing and a shrinkage of the lattice parameters. Using the (311) peak, it was identified that the lattice parameter of the reaction product was 0.6056 nm, while that of the reference was 0.6121 nm. Presumably the iodine efficiency discussed above can explain this lattice parameter change. In other words, the (Cu:I) ratio of (1.00 : 0.82) might contribute to the lattice parameter contraction.

The morphology of the reaction product was investigated using the SEM technique, and the results are shown in Fig. 5. It was observed that the interface between the copper mesh and CuI was clearly seperated, as shown in Fig. 5a. It is interesting to observe that the mesh-CuI interface region has a porous morphology while the outer surface of the CuI is composed of densely packed plate-type crystallites (Fig. 5b). The cross-section picture in Fig. 5c revealed that the CuI layer was grown up to a thickness of 24 μm without pores for iodine transport. These results suggest that the transport of iodine from the outer surface of CuI into the Cu-CuI interface was fast enough to generate a 24 μm thick layer within a 1.5 h reaction time. Here, it should be noted that the value of 24 μm is not the maximum thickness that CuI can grow before a spontaneous peeling-off. Further investigation should be done to determine the maximum CuI thickness including its effect on the Cu-I reaction rate.

4.Conclusions

The feasibility of a copper mesh as an iodine capturing medium during the pyroprocessing was investigated within a temperature of 100 ~ 400℃. The optimal reaction temperature was identified to be within the 300 ~ 400℃ range based on the iodine capturing efficiency. At 400℃, 6wt% of iodine was capture by a single copper mesh, and this promising result suggests that the stacking of copper meshes will be a reasonable approach for the complete capture of iodine generated from the pyroprocessing. Future works on the reaction rate between Cu and I, the complete consumption of a copper mesh, and the complete capturing of iodine might provide methods for the commercial application of a copper mesh as a cheap and effective medium for iodine capturing. In addition, it would be an excellent future research issue to identify the reaction between Cu and CsI, which is considered as one of main constituents of volatile iodine compounds.