Electrochemical Metallization Memories

In this chapter, we describe a resistance change technology that is based entirely on metal cation transport and redox reactions in simple two-terminal metal–electrolyte–metal/metal–insulator–metal (MEM/MIM) device structures 005B [1]. The device uses an oxidizable (soluble) active electrode, typically (but not limited to) Ag or Cu. The counter electrode is an inert (not soluble in the solid electrolyte) electrode, for example, W, TiN, and Pt. An oxide, chalcogenide, or halide material between the electrodes serves to transport the metal cations [2]. A positive voltage applied to the oxidizable electrode leads to the dissolution of the metal and deposition of a metallic (Agor Cu-containing) filament at the opposite electrode. This metallic filament ultimately bridges the relatively insulating ion conductor and defines the low-resistance on state (or states). The filament is dissolved by applying a voltage of the opposite polarity, which returns the cell to a high-resistance off state, defined by the area of the cell and the resistivity of the ion conductor. The basic elements of the device may be configured laterally or vertically. Whereas lateral devices, which typically have the electrodes placed at different points on the surface of a thin ion conducting film, may be utilized in a variety of applications, for example, radio frequency (RF) switches [3], it is the vertical configuration that is most applicable to memory devices. Vertical cells occupy the smallest possible area, which is critical for high-density memory arrays. These devices are nonvolatile, have the requisite low power consumption, and possess excellent prospects for scalability to atomic dimensions [4]. In addition, the distance between the electrodes is defined by the thickness of the ion conducting layer and can therefore be very small, for example, a few nanometers to a few tens of nanometers. This nanoscale distance allows rapid bridging by the electrodeposit, which results in fast switching. An example of the operation of a vertical cell is shown in Figure 17.1. As with any novel technology that is being examined by several research groups simultaneously, various names and acronyms have been applied to this