GOALI: To Understand Crystallization and Amorphization Dynamics in Phase-Change Memories by Linking Electro-Thermal Models, Electrical Experiments and TEM Characterization

University Of Connecticut, Storrs CT

Investigators

Abstract

Nontechnical description: Phase-change memory, an electronic memory concept that utilizes materials whose electrical resistance can be rapidly and repeatedly switched between high and low magnitudes, holds great promise for more efficient computation and data storage. This collaborative project between the University of Connecticut and IBM Watson Research Center combines computation, electrical characterization and microscopy to gain a better understanding of several phase-change materials phenomena that are critical for data storage devices. Results from this work can open new possibilities for future materials synthesis and electronic devices. This project supports the education and training of a diverse group of students in Electrical Engineering and Materials Science and Engineering, and also the broader community through several outreach programs. The students and post-doctoral researchers directly involved with the project benefit from access to state-of-the-art facilities and opportunities for first-hand contributions to fundamental studies with direct technological applications. Technical description: Large-scale use of phase-change memory has been hindered by the high power required to heat the nano-scale element above crystallization or melting (for melt-quench amorphization) and by limited reliability due to resistance drifts of the metastable phases, elemental segregation, and void formation upon extensive cycling. If these device characteristics can be improved, phase-change memory can be integrated on top of conventional silicon electronics as high capacity, non-volatile on-chip storage, leading to significant performance and energy improvements. This project focuses on the crystalline-amorphous transition, including the possibility of solid-state or electronic amorphization which can lead to lower power devices. In addition, it investigates correlations between electronic properties, such as resistance drifts, and changes in microstructure. Finally, the team is also addressing details of the lesser known cubic-hexagonal transition that is not typically used but may be advantageous in some cases. This work requires development of techniques for in-situ transmission electron microscopy experiments, suitable electrical characterization techniques and improved models that combined can provide a better understanding of phase transition dynamics across different time and temperature scales.