Recent advances in carbon capture create new opportunities for recycling CO₂ into liquid fuels to store intermittent electrical energy¹. One option is to co-electrolyze CO₂ and H₂O at high temperature using solid oxide electrolysis cells (SOECs). Although promising, the factors controlling rates of CO₂ and H₂O reduction in SOECs are not well understood, hampering development^2,3. Traditional electrochemical techniques have difficulty resolving the mechanisms and kinetic parameters of the individual steps governing CO₂ and H₂O reduction. This limitation can potentially be overcome using linear and non-linear electrochemical impedance spectroscopy (EIS and NLEIS) in conjunction with dynamic measurement of gas-phase species using differential electrochemical mass spectrometry (DEMS).

Regarding co-electrolysis, mixed ionic electronic conductors (MIEC) have gained interest as alternatives to nickel-yttria stabilized zirconia (Ni-YSZ) because the active region is not limited to the triple phase boundary and they are more stable in reducing environments³. The MIEC studied here, gadolinia-doped ceria (GDC), has been well characterized as an electrolyte³ and is a promising cathode in SOECs. Unlike Ni-YSZ, it does not coke in carbon environments or oxidize completely in a feed of water and CO₂.

We have performed NLEIS and EIS measurements of co-electrolysis on button cells composed of GDC as the working electrode with electrolyte YSZ under various temperatures and gas compositions. Gas atmospheres surrounding the cells contain mixtures of water, CO₂, CO, H₂, and carrier gas such as Ar or N₂, used to manipulate the oxygen partial pressure of the system and thus the oxygen vacancy concentration of the surface and bulk electrode. The data we have collected on co-electrolysis of CO₂ and water on GDC have been compared with model NLEIS spectra developed to include CO­₂ reduction, water reduction, and reverse water gas shift reaction mechanisms and rate determining steps. A rough vacuum and low volume reactor was developed to cross-check the proposed mechanisms through measuring a gas phase response in DEMS and a micro-kinetic model was developed to predict product distributions corresponding to the mechanism that most closely matched the results.