First-principles modeling of alumina and aluminates high-k dielectrics for non-volatile memory applications
(2012)
Files
NEW-Thesis-K-SANKARAN-19-09-2012.pdf
Restricted Access - Adobe PDF
- 11.76 MB
Details
- Authors
- Supervisors
- Rignasese, Gian-Marco
- Abstract
- (en) Portable electronic devices call for reliable storage of increasingly large amounts of data. In this framework, the development and optimization of solid state non-volatile memories is crucial for further progress. The predominant non-volatile memory is based on the NAND Flash technology, i.e. Floating Gate. So far, its basic architecture has remained unchanged since its first introduction. This era of “happy scaling” of the physical dimensions is now coming to an end, as Floating Gate devices cannot shrink further due to the severe physical limitations. Two different routes are being followed. On the one hand, the introduction of high-k/metal gate structures in place of traditional SiO2/poly-Si stacks, as it happened for logic, has been proposed for Flash memory. This replacement together with charge trapping medium concept (i.e, the TANOS cells) is considered to be a viable alternative candidate for scaling the NAND Flash technology below the 30-nm node. Naturally, these “higher-k” oxides ought to offer the same properties in terms of stability, controllability, data retention and low defect concentration as SiO2. For this transition to occur, it is hence of prime importance to get a detailed understanding of the properties of the possible material candidates and of their related defects. On the other hand, extensive research efforts are devoted to develop a successor of the Flash technology and to drive the scaling limits beyond the 10-nm node. Among the possible alternatives to Flash technology, resistive switching based memories are currently the object of intense investigation. More particularly, the conductive bridging random access memory (CBRAM) is a very attractive concept. This type of memory could revolutionize the current memory market. By combining the speed of static random access memory, the density of dynamic random access memory and the non-volatility of Flash memory, this emerging technology has the potential to merge the benefits into a single standalone product for future ICT applications. However, this technology still needs maturing in terms of reliability, performance, and production costs to compete with the present technologies. A thorough investigation of the physical mechanisms on which the device relies has to be performed. In order to overcome these concerns, first-principles (ab initio) modeling techniques are choice tools to provide insights into the very nature of these materials, their associated defects the high-k dielectrics in TANOS cells and the detailed driving mechanism of the conduction bridge-formation in CBRAM.
- Affiliations
Citations
Sankaran, K. (2012). First-principles modeling of alumina and aluminates high-k dielectrics for non-volatile memory applications. https://hdl.handle.net/2078.5/160809