Non-volatile memory is critical for all aspects of modern computing, as well as for next generation of digital technologies like the Internet of Things and neuromorphic computing. Among non-volatile memory technologies, resistive random access memory based on metal oxide films as resistive switching layers has the potential for high-speed, low operation voltage, low power consumption, and good endurance properties that enable the highest performance at the lowest cost. However, metal oxide resistive random access memory also faces some critical challenges such as the unpredictable forming process. Another challenge is the variable resistive states from one film to another and from one point to another across each film. The collaborative project between the US team (Univ. at Buffalo and Purdue) and the UK team (Univ. of Cambridge) will develop a highly innovative, scalable, and advanced materials technology to overcome the current technical limitations of emerging resistive memory. The materials platform is HfO2, a widely used material in the semiconductor industry. Unlike previous work on this material, the current project will precisely engineer HfO2 microstructures in new ways to create highly controlled switching properties. The broader technological impacts are built on established industry collaborations. The research program is well integrated with education and outreach programs at all three campuses, including: 1) training young researchers with multidisciplinary research skills in an international research environment; 2) implementing resistive memory concepts in materials science and engineering curricula through teaching; 3) disseminating research findings to broader audiences through outreach programs.
While commonly-used metal/metal oxide/metal structures for resistive random access memory have conduction filaments that are nucleated randomly, the design in this project incorporates engineered vertical interfaces in either vertically aligned nanocomposite or fine-grained columnar structures to guide the conduction channels. These pre-defined interfaces enable the formation of precise and non-random vertical conducting paths with high densities for high performance resistive random access memory, without the need for a high voltage forming process. This project advances knowledge by combining well-integrated capabilities to synthesize, characterize, design, and fabricate resistive random access memory devices with targeted properties and performance. Specifically, it will translate the ideal engineered materials systems which has been already demonstrated by this team in epitaxial nanocomposites to simple binary oxides such as HfO2 on Si. These films will be initially grown by pulsed laser deposition. The knowledge learned from the films grown by pulsed laser deposition will be then implemented to industrial tools of sputtering and atomic layer deposition to achieve nanoengineered HfO2-based films with ~few nm sized columnar-grains. Finally, individual memristors and crossbar array structures will be fabricated, and the key parameters of the devices characterized. Furthermore, a set of unique characterization tools will be used to reveal the interplay between the device performance and the materials properties. The ultimate goal of the project is to develop a forming-free, highly uniform, high density, low power, high on/off ratio, superior endurance resistive memory through the formation of controlled oxygen vacancy concentration and perfect conducting channels in resistive switching metal oxide layers.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
