Bertrand Morel

Chemical and Electrochemical Stability of Copper in Molten KF-2HF

Fluorine is produced from electrolysis of KF/HF mixtures at around 95°C. In the cell configuration, carbon anodes are screwed onto a copper busbar. Much attention has been paid to the stability of copper since the corrosion and re-deposition of this metal on the cathodes is one of the main factors (side reaction) which limit the production yield of fluorine gas, the lifetime of the cells and the development of new electrolyzers. Therefore, in the frame of this study, various experiments were carried out to determine the corrosion rate of copper for a wide range of HF ratio and temperature. A statistics approach of the electrochemical data allowed us to predict the corrosion rate for many the operating conditions. At OCV, copper shows a good corrosion resistance even for high HF ratios. However, under a 6 V anodic polarization, copper corrosion rate raises drastically with an increase of the temperature and / or the HF ratio. Characterization techniques have shown that only a thin copper fluoride layer has been detected on copper at OCV. By contrast, two types of copper fluorides were evidenced at the electrode surface (CuF2 and KCuF3) when a potential was applied to Cu. Under anodic polarization, a thin CuF2 layer is formed at the copper surface whereas KCuF3 is detected at on the electrode surface resulting from the precipitation of Cu2+. For a better interpretation of the results, erosion-corrosion phenomenon issued from the fluorine bubbles impacts and electrolyte movements has been highlighted by weight losses of copper pieces. The breakdown of the passivation layer on copper and the exposition of the surface to the corrosive medium imply a quicker degradation of the metallic pieces.

Corrosion of iron in liquid uranium hexafluoride

ABSTRACT Here is reported a dedicated and reliable experimental setup which has been designed to study the corrosive effects of liquid uranium hexafluoride (UF6) on pure metallic iron at 80°C and 3 bars of pressure. Three corrosion tests have been carried out, ranging from several hundred to almost one thousand hours. No uniform corrosion was observed on early-stage samples. However, a nodule facies consisting in FeF2 and UFx (with 4 < x < 5) deposits is observed at the surface of the iron samples. These FeF2 nodules grow with time and merge, leading finally to the formation of an almost continuous FeF2 layer. The thickness of the layer over time seems to follow a parabolic kinetics law, finally reaching 15 µm after about one thousand hours of immersion in liquid UF6. After the nodules have merged, many cracks can be observed in the FeF2 layer, probably due to the growing stresses as the corrosion layer is formed.