Localized carbon deposition in solid oxide electrolysis cells studied by multiphysics modeling

Maria Navasa*, Henrik Lund Frandsen, Theis Løye Skafte, Bengt Sundén, Christopher Graves

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review


Solid oxide electrochemical cells (SOCs) can store electrical energy in the form of chemical fuels with high efficiency by electrolysis of CO2CO2 and H2OH2O. However, achieving commercially relevant lifetime is hindered by degradation mechanisms such as carbon deposition, which can even destroy the cell especially during electrolysis where carbon formation is electrochemically driven at the electrode-electrolyte interface. Here we used a three-dimensional multiphysics model to simulate a SOC performing CO2CO2 electrolysis and determine the operating conditions and locations in the porous nickel-based electrodes where carbon deposition is expected based on local conditions (gas composition, temperature and overpotential) crossing local thermodynamic thresholds. It is found that CO/CO2CO2 gas diffusion gradients and cooling from the endothermic electrolysis reaction are important driving forces for carbon deposition to occur locally when it is not expected based on the outlet CO concentration. Furthermore, correlation with a set of experimentally determined threshold operating points indicates that carbon deposition occurs primarily by the Boudouard reaction rather than by direct electrochemical reduction of CO or CO2CO2 to carbon for the studied cell type. Variation of fuel electrode porosity and thickness shows that these methods of reducing gas diffusion limitations widen the operating window that avoids carbon deposition.
Original languageEnglish
JournalJournal of Power Sources
Pages (from-to)102-113
Publication statusPublished - 2018


  • Carbon deposition
  • SOEC
  • Multiphysics modeling
  • Boudouard reaction
  • Thermodynamic threshold


Dive into the research topics of 'Localized carbon deposition in solid oxide electrolysis cells studied by multiphysics modeling'. Together they form a unique fingerprint.

Cite this