The overall goal of WP2 has been to explore and evaluate steel grades, coatings, and manufacturing processes to reduce the overall cost of the interconnects without affecting the performance. The highlights are:
• A successful demonstration of the feasibility of using SoA high-volume roll-to-roll manufacturing methods for interconnect.
• A successful long-term test of the coated interconnect chemical stability for up to 3,000 hours, where the coatings decrease the chromium evaporation by at least 30 times. Investigations on old samples of Sanergy HT 441 proved the stability of the coating for 38,000 h.
• Once coated with a protective coating the low-cost steels AISI 441 and K44M perform equally well as the specialized steel Crofer 22 APU with respect to corrosion rate, Cr evaporation and ASR measured in the lab.
• Demonstrated an ASR < 20 mΩcm2 at 850°C after 3000h of operation (M2.2).
In WP3 a new design of the interconnect with an optimized flow distribution was developed. This allowed for even thinner electrolyte supported cells to be introduced in the Sunfire stack design. The development was based on the very efficient 3D multiphysics model developed at DTU that considers couplings between flow, heat transfer, mechanical stresses and electrochemical reactions. FZJ investigated the corresponding flow simulations to ensure that the pressure drop will be maintained low. The interconnects were produced by hydro-forming by Borit and implemented in stacks by Sunfire. Furthermore it was shown that the pre-coated interconnects from Sandvik could be shaped by hydroforming, and if small cracks occurred during the forming these were closed after the heat treatment of stack assembly.
In WP4 novel contact layers were developed by DTU, which relied on in-situ reactive bonding, achieving well conducting and most importantly high mechanical robustness. Several iterations of contact layers were pursued. The main idea was using metallic powders as precursors to the spinel contact layers. When heated in the stack assembly process the metal powders would oxidise and react with adjacent layers (interconnect, cell) and form a strong bond. The initial formulation was CoMn and CuMn. The latter showed the best mechanical performance (more than 15 times stronger than state-of-the-art contact layers of sintered ceramic perovskite powders). It did however turn out that the Cu was attracting Cr from the interconnect steel, generating a pathway for faster Cr evaporation. Co did not have this challenge, but reacted / sintered less with adjacent surfaces. Thus combinations of Co, Cu and Mn was thus pursued. The latest generation showed 15 times higher robustness than state-of-the-art, while also not attracting Cr and vaporising Cr (showed by Chalmers University). Up-scaled printing of the contact layer with commercial printers was successfully carried out at Tecno Italia, and the printed interconnects were sent to the stack manufacturer. The robustness of the print could be improved.
In WP5 four different stack designs with different materials were produced to demonstrate the developed materials mentioned in the previous 3 WPs. A stack was successfully built using Sanergy® 441 HT as interconnect steel/coating solution and tested for 3,500 h at 800-850 C. The Sanergy® HT 441 with CeCo-coating showed higher degradation of ASR compared to Crofer with MCF coating, but performed better at lower temperature (750 C). The stack with the new contact layers had similar performance as the standard solutions at the respective companies without any optimisation of the processing, which is thus a promising result.
In WP6 the technical improvements from the other WPs was converted into monetary numbers to prove industrial relevance and applicability by AVL. It was shown that the mass manfacturing routes would be commercially competitive as compared to in-house production despite increased overheads from more manfacturers. The reason was the scalable processes of roll-to-roll and high speed printing. It was shown that the interconnects cost could be reduced to below 5 € per interconnect.
The work was disseminated in WP7 - through 12 published papers and 4 more in preparation, 11 conference presentations, and two workshops with 32 participants in each - from both academia and most of the SOC stack manufacturers in Europe. The works have thus been well disseminated or will soon be. All papers are open access.
