A Decade of Improvements for Solid Oxide Electrolysis Cells. Long-Term Degradation Rate from 40%/Kh to 0.4 % Kh

Abstract

Solid oxide electrolysis cells (SOEC) have the potential for efficient large-scale conversion from electrical energy to chemical energy stored in fuels, such as hydrogen or synthetic hydrocarbon fuels by use of well-known catalysis processes. Key issues for the break-through of this technology are to provide inexpensive, reliable, high performing and long-term stable SOEC for stack and system applications. At DTU Energy (formerly Department of Fuel Cells and Solid State Chemistry, Risø National Laboratory), research within SOEC for more than a decade has led to long-term degradation rates on cell level being improved from 40 %/kh to 0.4 %/kh for tests at -1 A/cm2 (figure 1). In this paper, we review the key findings and highlight different performance and durability limiting factors that have been discovered, analyzed and addressed over the years to reach the tremendous increase in long-term stability for SOEC as illustrated by the cell tests in figure 1. First, some of the early SOEC tests showed that even though solid oxide cells (SOCs) are in principle reversible cells, long-term degradation trends are not identical and especially their sensitivity towards different impurities varies with choice of operation mode. Glassy phase impurities e.g. from sealing material and even ppm-level of impurities in the gas supply can cause significant performance loss; however such effects can also be minimized by proper choice of set-up and gas cleaning designed for SOC. Second, operating cells that have LSM/YSZ oxygen electrodes at high current densities showed detrimental ohmic resistance increase due to oxygen bubble formation in the electrolyte part closest to the oxygen electrode. This effect is related to the oxygen-electrode overpotential and by introducing higher performance, mixed ionic-electronic conducting (MIEC) oxygen electrodes such as LSC/CGO based electrodes not only higher initial performance but also improved long-term stability was obtained. Introducing such cobaltite electrodes also introduces the need for a barrier layer between the YSZ electrolyte and the oxygen electrode, and a barrier layer optimized for fuel cell operation is not necessarily optimal for electrolysis operation of the SOC. Third, when having minimized the effect of impurities and applied a high performance MIEC oxygen electrode and a barrier layer designed for electrolysis operation, the largest contribution to the cell resistance for SOEC now originates from the Ni/YSZ fuel electrode. Ni/YSZ electrodes are still attractive in a technological perspective due to rather inexpensive and up-scalable production of SOEC and our recent work show a wide “processing window” e.g. with respect to solvents, multilayer tape casting process, sintering temperature, Ni/YSZ ratios etc. Furthermore, increased detailed knowledge on the correlation between microstructures (3D reconstructions) and electrochemical performance (impedance spectroscopy) has laid the ground for significant improvements in both initial electrochemical performance and long-term stability for SOEC operated at high current density. These latest and significant improvements were obtained by electrode microstructure optimization and this illustrates the importance of detailed knowledge on the complex interplay between processing parameters, obtained microstructures and resulting electrochemical performance and durability. Finally, what are the next steps for further SOEC improvements? We’ll here present “appetizers” for our ongoing electrode research and development work e.g. on oxygen electrodes based on infiltration of electro-catalysts in oxide ion conducting scaffolds and microstructure and composition wise improved oxide ion conductors for the fuel electrode part of the SOEC where demands for long-term stability and mechanical robustness are also considered. Figure 1 : Development of cell voltage for solid oxide electrolysis cells for steam electrolysis at – 1 A/cm2 at 850 °C and 800 °C, respectively for single cell test performed in 2005 and 2015 at Risø National Lab./DTU Energy. Long-term voltage degradation rates in the linear part of the cell voltage curves were 40 %/kh and ~ 0.4 %/kh, respectively, for SOEC test performed in 2005 and 2015. Figure 1