THE SCHOOL OF MATERIALS SCIENCE AND ENGINEERING

GEORGIA INSTITUTE OF TECHNOLOGY

Under the provisions of the regulations for the degree

DOCTOR OF PHILOSOPHY

on Friday, May 27, 2022

11:00 AM

in Love 109

& via

BlueJeans Video Conferencing

https://bluejeans.com/252222750/0902

will be held the

DISSERTATION DEFENSE

for

Jinho Hah

"Investigation of Processing Condition on the Reliability and Performance of Future Memory Devices"

Committee Members:

Prof. Samuel Graham, Advisor, ME/MSE

Prof. Eric Vogel, Co-Advisor, MSE

Prof. C. P. Wong, MSE

Prof. Alan Doolittle, ECE

Prof. Arijit Raychowdhury, ECE

Abstract:

About a quarter of the total semiconductor market is comprised of memory technologies. However, due to the limits in scalability related to Moore’s Law, the need for additional memory capacity and bandwidth is increasing. Currently, dynamic random-access memory (DRAM) is in high-volume but is reaching its scaling limit. To overcome the DRAM scaling limit with higher bandwidth, lower power, and smaller surface area, through-silicon-via (TSV) stacking method via thermo-compression bonding (TCB) technology has been widely employed due to the placement accuracy for 3D packaging technology and manufacturing 3D high bandwidth memory (HBM) DRAM modules. In addition to utilizing TCB technology to DRAM modules, memristors, as an emerging memory is another memory class that can help overcome the scaling limit and holds a huge promise to overcome the current bottlenecks.  

Two project areas will be discussed. In the first project, comprehensive and comparative analysis between the TCB-processed and reflow-processed solder joint reliability performance is investigated. This study investigates the failure analysis of the solder joints via detailed solder joint and solder joint/bond pad interface characterization to understand the reliability issues of the emerging TCB-processed packages. In the second project, memristor, a class of resistive random-access memory (RRAM) will be explored. The RRAM or memristor is an emerging memory that shows promise to overcome the current obstacles in conventional memory systems based on a von-Neumann architecture. So far, the difficulty in achieving multiple resistance states and obtaining resistance linearity remain as the challenges towards commercialization and for in-memory computing and synaptic device (i.e., neuromorphic) applications. In this work, the effect of the chemical environment of the HfOx/Ti interface and the processing condition during analog operation on the memristor performance are explored. The presence of excess oxygen concentration at the interface and fundamental understanding of the local thermal and chemical environment at the memristor filament region during analog RESET operation are investigated to better understand the memristors’ resistive switching mechanisms.