Recent Advances on the Modeling of Phase Change Materials and Devices
- PDF / 784,735 Bytes
- 12 Pages / 612 x 792 pts (letter) Page_size
- 13 Downloads / 143 Views
1072-G06-05
Recent Advances on the Modeling of Phase Change Materials and Devices Andrea Leonardo Lacaita1,2, Ugo Russo1, and Daniele Ielmini1 1 Dipartimento di elettronica e informazione, Politecnico di Milano and IU.NET, Piazza L. Da Vinci, 33, Milano, 20133, Italy 2 IFN-CNR, Milano, 20133, Italy ABSTRACT As non-volatile memory technology is approaching the 45nm generation node and conventional Flash technology is facing severe scaling limitations, phase-change memory (PCM) is gaining momentum as a reference emerging memory. However, much development is still needed at technology, device and material levels, which in turn requires a significant improvement of the physical understanding of programming, reliability mechanisms and scalability of the new technology. This work reviews the most recent advances on modeling concerning PCM, spanning from program and read operations to device reliability. Electro-thermal simulations are employed to address the reduction of programming current for the reset operation (i.e., the transition to the high resistive amorphous state), and the trade-off with readout performances will be discussed and analyzed. Then, reliability issues concerning the stability of the programmed amorphous phase will be investigated, with particular interest on the spontaneous crystallization leading to data-loss. Starting from electrical retention measurements, an analytical model for nucleation and growth in the amorphous phase will allow to draw guidelines for material engineering and reliability improvement, which will be addressed on the basis of reported experimental results on different chalcogenide materials.
INTRODUCTION In the field of non volatile memories, phase change memory (PCM) has shown to be a promising emerging technology in the last few years, long after the original proposal in the seminal paper by Ovshinsky [1]. Indeed, PCM extended cycling endurance, good retention and easy integration has motivated large interest from several semiconductor industries and several test demonstrators have been developed for PCM, which nowadays appears as the most promising memory concept for replacing Flash memory in the near future [2-4]. In PCM, data are stored in the crystalline or amorphous phase of an active chalcogenide material. This can be achieved by heating up the material with external pulses, in order either to melt and quench the material in the amorphous phase (which provides the high resistive state), or to accelerate its transition to the crystalline phase (which provides the low resistive state). Nowadays, positive performances have been demonstrated, nevertheless several critical and open issues remain in the understanding of the reliability and scaling limitation of this technology.
One major concern regards the reset programming operation, where an electrical pulse is applied to the cell in order to obtain a high resistive, or reset, state. Indeed, reset programming requires melting and fast quenching of the active material, hence large temperatures are to be obtained by Jou
Data Loading...
