Hydralloy® C5, an intermetallic TiMn2-based alloy, has been manufactured industrially (GfE, Nuremberg) for decades and is used on a large scale for hydrogen storage. During use, the alloy is stored in gas-tight and pressure-resistant storage containers. At the end of service, the alloy is a fine powder with pyrophoric character (Ti- and Zr- content). This significantly hinders the safe extraction from the containers and subsequent recycling of the alloy due to unavoidable reactions with ambient air. The major concern on passivation and maximum permissible content with O/N must be clarified for safe handling in ambient air as well as regarding the pyrometallurgical recycling. Considering this, and in preparation for the opening of real large-scale storage containers, end-of-life Hydralloy C5 was synthesized with two different levels of O (~0.15 and ~1 wt.%) and N (~0.04 and ~8 wt.%) contamination. Vacuum induction melting (VIM) and cold crucible arc melting (CCAM) were chosen as potentially suitable for recycling. The preliminary remelting trials from both aggregates ascertained that the recovery of metal content is not feasible with heavily O/N-contaminated alloys. It is concluded that extreme caution should be taken to minimize contamination when extracting the powdered alloy from the storage containers. Hydralloy C5 with moderate gas impurities (~0.15 wt.% O and ~0.04 wt.% N) can be remelted, on the other hand, in both VIM and CCAM. Contact between molten Hydralloy C5 with selected refractories (Al2O3-TiO2 and CaO-stabilized ZrO2) in the VIM leads to the formation of a multi-layered transition zone dominated by Ti and Zr. While the Al2O3 in the titanium aluminate is infiltrated and reduced by Ti and Zr, the crucible wall made of CaO-stabilized ZrO2 remains intact. Despite low gas contents, significant losses in melt yield are recognized due to crucible wall deposits from the formation of non-metallic inclusions during VIM. Against this background, the use of fluxes is being considered for future melts in addition to the use of deoxidants.
Krusenbaum et al. (Tue,) studied this question.