The rate of oxygen exchange across mixed conducting oxide surfaces, often described via measurements of the surface exchange coefficient k, is a key metric impacting high temperature electrochemical device efficiency. Continuous measurement over time of native surface k values in realistic operating conditions can be accomplished by the contact-free optical transmission relaxation (OTR) approach, which relies on changes in oxygen stoichiometry being proportional to changes in optical absorption that can be detected by in situ optical transmission measurements. Conventional approaches to measuring k may suffer from being non-continuous (e.g., tracer exchange with ex situ secondary ion mass spectrometry), and therefore unable to observe changes in k over time, or require metal current collectors (e.g., ac-impedance spectroscopy measurements of area-specific resistance, or electrical conductivity relaxation), which could modify the surface reaction mechanisms and/or kinetics being probed. It is therefore of interest to compare k values obtained by ac-impedance spectroscopy (AC-IS) using metal current collectors and those obtained by OTR on native surfaces to examine the impact of current collectors and different measurement approaches. While k values obtained by various methods are already available in the literature, comparisons of such values measured on different samples may not provide much insight, as k can vary orders of magnitude depending on surface chemical composition, which is influenced by measurement conditions, processing, and thermal history. Therefore, in the present study, the two techniques were compared simultaneously on the same sample.

A thin film of the mixed conductor SrTi_0.65Fe_0.35O_3-x (STF35) was prepared by pulsed laser deposition on a transparent, ionically conducting (Zr,Y)O_2-δ single crystal substrate. Its k values were measured by AC-IS on one region and by OTR on another region of the film, at 600 °C as a function of oxygen activity, controlled by gas partial pressure and/or dc bias applied across the substrate. Upon changing the oxygen activity, optical transmission relaxations resulted, corresponding to changes in the film oxygen stoichiometry and corresponding concentration of optically absorbing oxidized Fe (~Fe⁴⁺) species; such relaxation profiles were successfully described by the equation for surface exchange-limited kinetics appropriate for the film geometry, to obtain k_chem values. From AC-IS, the fitted resistance and capacitance of the low frequency arc were used to determine both k_chem (from the time constant) and k_q (from the resistance), and additionally k_chem values were estimated by multiplying k_q by the thermodynamic factor (γ).

Thin film γ values, determined from the film’s apparent volumetric chemical capacitance, were typically higher than the bulk values, determined from thermogravimetric measurements of non-stoichiometry, and varied with current collector material used (porous Au vs. porous Pt), suggesting a possible interfacial capacitance contribution. The AC-IS-derived k_chem values and k_q values multiplied by the thermodynamic factor (both bulk and thin film) were consistently orders of magnitude higher than the k_chem values obtained by OTR, regardless of current collector material (Au or Pt). The results suggest a possible enhancement in k by the metal current collectors, which has similarly been observed previously for films of the mixed conductor (Pr,Ce)O_2-δ. Additionally, long-term degradation in k_chem and k_q values obtained by AC-IS was also attributed to deterioration of the current collector, while hardly any degradation was observed in that time period in the optically-derived k_chem values. The results suggest that, while the current collector could modify surface exchange coefficient evaluation by AC-IS, the contact-free OTR method offers a continuous, in situ approach to evaluate native surface exchange behavior.