MFMISFET based memory cell

Period of Performance: 07/24/2003 - 01/24/2004

$69.8K

Phase 1 SBIR

Recipient Firm

Cova Technologies, Inc.
5061 North 30th Street, Suite 105
Colorado Springs, CO 80919
Principal Investigator

Abstract

Nonvolatile semiconductor memory is critical in many DOD applications - particularly when radiation hardening is required. Ferroelectric memory technology offers superior performance, notably better write-endurance and write-speed, lower power consumption and it is not subject to the scaling limitations of other technologies. We propose to develop a ferroelectric memory cell based on the MFMISFET device using both metal layers as control terminals. While the MFMISFET technology is somewhat more complex than MFISFET technology, the innovative design proposed will result in a much more reliable memory cell. In Phase I, COVA Technologies will develop a simple experimental circuit to validate the fundamental technology. A comprehensive circuit model for the MFMISFET will be developed and validated using test data from the experimental circuit. The model will also be used to simulate several memory array architectures with MFMISFET cells to validate reliable programmability and readability. In Phase II we will build a test chip. We will fully characterize these devices and determine the write/read performances, endurance, retention and particularly the disturb limits of the proposed cells. In Phase III we will commercialize the technology with the assistance of a Strategic Partner by designing and manufacturing high-density ferroelectric memories based on the MFMISFET cell. The nonvolatile semiconductor memory market is very large (~$10 billion), currently dominated by flash memory. However, it is recognized in the industry that flash will run out of steam in about two technology generations. Ferroelectric memories are one of the most attractive next generation technologies. The proposed ferroelectric technology with one transistor as the only element and a MFMIS structure (metal-ferroelectric -metal-insulator-silicon) has the potential of penetrating the flash memory market since it will preserve the cost advantages of a product with a small cell size based on a single ferroelectric transistor, but increase the robustness and reliability of the memory cell with the addition of an extra electrode. The proposed technology is inherently radiation hard, an important attribute for military applications.