Publication date: 10th April 2024
Ion-conducting metal oxides are ubiquitous in various electrochemical energy conversion applications, offering high conversion efficiencies and environmental stability. The ionic transport properties of polycrystalline electrolytes used in practical electrochemical devices, however, can be dominated by the internal grain boundary (GB) interfaces. The blocking of ionic transport at GBs has typically been assigned to local space charge barriers that form due to the segregation of lattice defects or impurities to the grain boundaries, resulting in the development of net charge at these interfaces. These space charges induce mobile ion defect depletion zones adjacent to the GBs that can decrease the overall ionic conductivity by orders of magnitude and increase their corresponding activation energies. While this space charge phenomenon associated with GBs in solid electrolytes has been well known and characterized for some time, little effort to date has been made to demonstrate the ability to engineer these space charge barriers by control of the GB chemistry and/or structure. In recent years, our research team has demonstrated the ability to modulate ionic conduction in polycrystalline Gd doped ceria (GDC) solid electrolytes by illumination with above bandgap light[i]. The electrons and holes thus generated are separated by the built-in space charge fields localized at the GBs, leading to a net decrease in barrier height and thereby an increased effective ionic conductivity. In an attempt to optimize this opto-ionic effect, we have investigated means for controlling the built-in GB space charge barrier heights.
Pursuing an approach that we first demonstrated in Pulse Laser Deposited (PLD) ceria thin films in 2009,[ii] GDC films with 3 at % Gd were grown on single crystal substrates that served as sources of selective cation up-diffusion along the GDC GBs, taking advantage of the many orders of magnitude higher diffusivities of cations along oxide GBs compared to those in the grains. MgO and Al2O3 substrates were selected with the expectation based on the difference in ionic radius of cations from Ce that while Mg would substitute on Ce GB sites, leading to localized negatively charged species, Al would enter interstitially as a localized positively charged species. Given that such GDC materials, as grown, exhibit net positive GB charge,[iii] we expected that Mg in-diffused films would show decreased GB barrier heights while Al in-diffused films would show increased GB barrier heights. Indeed, we demonstrated this capability, dropping barrier heights to as low as 0.01 eV for Mg in-diffused films while increasing barrier heights to as high as 0.43 eV for Al in-diffused films. In this report, we review the in-diffusion experiments, discuss the derived Mg and Al GB diffusivities as determined from secondary ion mass spectrometry (SIMS) measurements and subsequent profile analysis, and briefly discuss the implication this may have for detecting illumination sources.