Project abstract:

Phase Change Random Access Memories (PC-RAM) are a type of non volatile memories using phase change materials. In fact, these materials are so called because their crystallographic structure can be switched between an amorphous state and a crystalline state, where the atoms are respectively randomly disordered in the space or ordered on a periodic lattice. These materials present a large difference of more than 2 orders of magnitude in electrical resistivity between the amorphous semi-conducting phase and the crystalline semi-metallic phase (Fig 1). Consequently, it is possible to use these materials for non-volatile random access memory devices where the read signal arises from the resistivity difference between the two phases. For writing, current pulses are applied to the material in order to induce switching through Joule heating effect.

Fig. 1. Variation of the resistivity as a function of the temperature during annealing of an initially amorphous GeSbTe sample. (Source : RWTH Aachen)

PC-RAM devices with very attractive specifications have been recently demonstrated so that a wide range of applications is now foreseen. Possible applications include for example Flash memories, DRAM, embedded memories and radiation-hard applications, addressing today a huge market of more than 100 billion US$ total annual value. Besides, this type of low power consumption memories is particularly well-suited for portable applications whose development is required for a user-friendly society. What makes this type of memories even more attractive is their capability of multi-level recording and their scaling ability, since this allows to develop very high density memories. Indeed, while most actual memory concepts are expected to fail down when scaling down, the performances of the PC-RAM are expected to improve. The industrialists are currently studying alternative solutions to overcome the difficult challenges faced by the current technologies below the 65nm technology node and they show a great interest for the PC-RAM technology. The research on the PC-RAM devices is currently led by American Companies, namely Intel and Ovonyx from which ST-Microelectronics recently bought the license.

In this framework, the objective of the PC-RAM project is to study the potentiality of this concept for future applications in nanoelectronics. This will allow us to draw the technical roadmap for the coming years and to acquire the knowledge necessary for future development of this technology by European Companies.

To achieve these objectives, stand-alone structures where the phase change material is sandwiched between two conductors (Fig. 2) will be fabricated and tested. Particular attention will be given to the scaling ability and not only on a theoretical point of view : e-beam lithography will be used to realise nanometer-sized devices. Besides, simulations will be performed to forecast the physical limitations of this memory concept. Finally, the integration of these structures into the complete memory device will be evaluated.

Fig. 2. Schematic drawing of stand-alone structures

The following milestones have been identified : choice of the PC material, results of electrical characterisation of the PC material (IV curve), report of the reading and writing conditions, evaluation of the writing performances (writing time, cyclability, multi-level capability), prototype of stand-alone structures, nanometer-sized switching elements, technical roadmap of the PC-RAM memories.