Alternative Perovskite Materials for ReRAM Memories: Understanding and tailoring resistive switching – Alps_Memories

Among the various emerging devices expected to replace conventional Flash memories, Resistive Random Access Memories (ReRAM) are currently attracting a strong scientific and industrial interest. Their operations are based on the switching between a low resistive and a high resistive state, which represents the two binary states. ReRAM devices have already demonstrated significant advantages over other technological options, such as high scalability, fast switching speed combined with low switching energy, low power consumption, strong endurance and data retention larger than ten years. Different types of ReRAM have been demonstrated so far: some of them exploit the breakdown properties of metal-oxides (Metal Oxide - Bipolar Filamentary, MO-BF), while other use the formation of a conductive bridge (CBRAM). The switching mechanisms in these devices are mainly based on the creation/annihilation of a conductive filament, formed, either by a path of oxygen vacancies, or by the oxidoreduction of metal such as silver or copper.

Perovskites based on transition metal-oxides also feature this kind of resistive switching behavior. The evaluation of the potential of this types of materials for future application as memory devices has hence to be carried out. A better understanding of the physico-chemical mechanisms involved in the resistive switching (RS) is therefore strongly required. The acquisition of this knowledge will grant a greater mastering of the fabrication and optimization of ReRAM devices, which currently exhibits a variability of performances which is too large to be exploited at an industrial scale.

This project focused on basic research will hence study the RS in perovskite materials with the aim of answering fundamental-science questions. It will allow to move from fundamental studies of functional oxides, to their direct application in emerging non-volatile memory devices by:

• Better understanding the nanoscale mechanisms governing the RS, charge carriers and interface effects. This will be achieved by providing for the first time a unique set of complementary physical and chemical cutting-edge characterization methods (in-operando and chemistry at interfaces) which have never been combined together.
• Demonstrating the effectiveness and application of the optimized materials in reliable perovskite-based ReRAM memories with appropriate performance for non-volatile applications, i.e. fast switching speed, large programming window, high endurance and retention. This will be accomplished by tuning of the RS properties by composition variation (cationic substitution and oxygen vacancy content), morphology control and epitaxial strain engineering.
• Project the viability of future perovskite-based memories by modelling the intrinsic device performance, allowing a benchmark with other technological options.