Electrochemical control has emerged over the last decade as a powerful approach for developing reprogrammable microelectronic devices. Such oxide based devices are characterized by remarkably large changes in chemical composition at low energetic cost. Both from a fundamental science and applications-oriented perspective, the ability to control the functional properties of metal oxides dynamically through their chemical composition opens a broad perspective in developing new classes of microelectronic devices. While such concepts have been demonstrated, important questions remain about how to optimize their device performance best. Information regarding ionic transport within the oxides remains elusive given the challenge of isolating it from the dominant electronic conduction. Moreover, many current device concepts not only rely on the transport of ions within a single oxide layer but across interfaces with additional layers as well. Adequate methods for studying the ionic properties in these ultra-thin films and isolating their transport kinetics are lacking. Their development will be vital for mastering overall device performance, such as speed, retention, and predictability.

To address these challenges, we have been investigating a model system capable of reversible ion transport and exchange under an applied field between two adjacent thin film solid oxide layers. Our investigations have focused on studying a Pr_xCe_1-xO₂/La_2-xCe_xCuO₄ bilayer stack that utilizes model materials for which the defect chemical and transport properties have previously been carefully characterized, that offer high ionic mobilities and can accommodate large levels of non-stoichiometry [1,2]. Building on our previous observations that we could reversibly modulate the resistance of the stack by over an order of magnitude through the transfer of ionic species, we discuss how our conductivity results can be interpreted via defect chemical models and show how such transfer can be tracked in situ inside a TEM by monitoring local valence and lattice dimension changes. Furthermore, we demonstrate that we can isolate the ionic transport kinetics, both within and across the solid oxide interfaces, by a combination of DC bias impedance measurement and dynamic current-voltage studies. This allows us, for example, to systematically study how the ionic mobility varies in Pr_xCe_1-xO₂ as a function of stoichiometry, as controlled by the degree of ionic titration. The findings in this work can be expected to aid in developing material selection and design criteria for similar bilayer systems and be used to achieve faster resistance switching speeds, larger resistance switching ranges, and longer device retentions.

[1] Tuller, H. L., Bishop, S. R., Chen, D., Kuru, Y., Kim, J.J. & Stefanik, T. S. Praseodymium Doped Ceria: Model Mixed Ionic Electronic Conductor with Coupled Electrical, Optical, Mechanical and Chemical Properties. Solid State Ion 225, 194–197 (2012).

[2] Kim, C. S. & Tuller, H. L. Fine-Tuning of Oxygen Vacancy and Interstitial Concentrations in La_1.85Ce_0.15CuO_4+Δ by Electrical Bias. Solid State Ion 320, 233–238 (2018).