What Has Been Achieved: The synthesis and measurement of epitaxial thin-film heterostructures containing high-entropy oxides with formulations tuned to support hopping conductivity that enables transient memristive behavior ideally suited for task-specific neural network systems. The material synthesis followed first principles predictions of equilibrium defect pairs that act as the charge hopping sites and that can be tuned by modifying the cation ratios of this 5-component base oxide. Pulsed measurement reveal transient conductivity that can be tuned by formulation and utilized in temporal data processing. These findings create a foundation for an expanded palette of IRG2 research activities.
Importance of the Achievement: Smoothly tunable hopping conductivity in memrsitor elements is highly desirable for neural network systems. Moreover, the conductivity transitions occur without physical filaments common in other actively investigated oxide systems that often lead to device-to-device non-uniformity, hysteresis, and drift.
How is the achievement related to the IRG, and how does it help it achieve its goals? One of the guiding high-entropy value propositions of IRG2 is that many-cation oxides with high solubility exhibit high crystalline fidelity while exhibiting a broad spectrum of local structures with great chemical diversity which is difficult or impossible to replicate in low-entropy systems. In the present case, these different local structures include electrically excitable defect states whose energy states, and thus transport, are composition tunable.
Where the findings are published: Efficient data processing using tunable entropy-stabilized oxide memristors, Nature Nanoelectronics, Sangmin Yoo, Sieun Chae, Tony Chiang, Matthew Webb, Tao Ma, Hanjong Paik, Yongmo Park, Logan Williams, Kazuki Nomoto, Huili G. Xing, Susan Trolier-McKinstry, Emmanouil Kioupakis, John T. Heron, and Wei D. Lu, DOI: 10.1038/s41928-024-01169-1