6 months
Cadre du stage/Internship framework:
The LETI environment
A unique scientific, industrial and cultural environment, with its research centers, university campus, 500 foreign companies and 40,000 scientists, engineers and technicians employed in the area, the Grenoble-Isère region, otherwise known as France’s Silicon Valley, mixes world-class intellectual and scientific dynamism with exceptional quality of life. It is the ideal springboard for Leti’s expansion. Located in the heart of a unique scientific, industrial and cultural environment, the CEA-Leti Institute for micro- and nanotechnology research offers researchers alike a rewarding place to work. You will grow in an environment where the scientific community is passionately engaged in technological research: men and women who are ready to share their expertise with you in your scientific and professional development.
In recent years the research on new data storage concepts has ramped up at impressive speed. While the memory market is demanding non volatile storage supports constantly with higher performances, the limits of the actual Flash technology are approaching. To meet the requirements of new applications and their higher capabilities, a new class of memory devices is under study, with the name of resistive memories.
Among those, CBRAM (Conductive bridge RAM) is particularly interesting since it conjugates low power specs (working voltages below 1V) with very high speed (in the order of hundreds on ns, down to a few ns for writing and erasing).
Fig 1: Schematics of a CBRAM working principle (Waser, Nanotechnology 22, 2011): Appling a positive voltage causes the build up of the Ag filament, which is brought back under reversed bias.
Fig. 2: Example of CBRAM integration in the BEOL (Back end of line) layer of a chip (Adesto).
The working principle of a CBRAM relies on the formation and disruption of a thin conductive metallic filament through a dielectric matrix. The memory cell is generally constituted by a chalcogenide (but several other materials are used too) sandwiched by an active and an inert electrode. The active one serves as reservoir of the metallic specie that forms the conductive bridge, which is in most cases Ag. Under the effect of the electric field Ag is transferred, via electro-migration, to the inert electrode, eventually shunting the electrodes. Reversing the bias causes the Ag to be back transferred to the active electrode, interrupting the conductive bridge. The information bit is stored in the resistive level of the cell. High resistance = 1, low resistance = 0 (see Fig. 1). One of the advantages of this technology is its relatively high simplicity in terms of cell structure. This kind of device can be easily embedded in the back end of line production process, strongly reducing manufacturing costs (see Fig. 2).
Travail demandé/Work description :
One of the most critic aspects of the technology is linked to data retention. The Ag filament naturally tends to dissolve in the chalcogenide matrix. This effect is temperature accelerated, since higher temperature causes higher Ag mobility and thus a faster loss of the programmed data. To improve data retention a more aggressive programming can be for instance exploited, leading to thicker Ag filaments and thus long lasting retention. However, since the microscopic Ag diffusion mechanism is still not completely understood, it is difficult to make predictions of the impact of programming on retention.
The candidate will investigate on this effect on data retention, aiming at the comprehension of the microscopic mechanisms that governs the filament dissolution. He/she will electrically characterize CBRAM cells under quasi-static and pulsed condition and will perform data retention tests. Moreover he/she will focus on more in-depth tests aimed to a deeper physics comprehension of the data loss mechanism, trying to correlate electrical measurements with physico-chemical data (for instance, Scanning Electron Microscopy, advanced Transmission Electron Microscopy, x-ray photoelectron spectroscopy or X-ray diffraction).
The candidate will also work toward a physics-based modelling, capable to reproduce experimental data and to make predictions on data retention performances.
The student will work in direct contact with several experts, both in the material engineering group as well as in the integration group and in the characterization task force. He will be involved in an extremely rich environment, participating in a project with industrial partners.
.
Unité d’accueil/Host Institution
Direction/Département/Service/Laboratoire
CEA-LETI – DRT/SCME/LCTE
Adresse postale/Postal address
CEA/GRENOBLE , MINATEC Campus, 17 rue des Martyrs
38054 Grenoble CEDEX 9 France
Responsable technique/ Technical Supervisor
Nom-prénom : CAGLI Carlo
Téléphone : 04 38 78 9057
EMail : carlo.cagli@cea.fr
Possibility of thesis
- Keys words:
Memory technologies
- Laboratory:LETI / DCOS
- CEA code:CAGLI_Carlo
- Contact:
carlo.cagli@cea.fr