An explosion in global data production is observed today, largely attributed to the emergence of 5G and the Internet of Things. Faced with the challenges of managing vast amounts of data and to avoid a significant increase in global electricity consumption due to data storage, it is necessary to develop efficient, energy-efficient, and dense non-volatile memory technologies.

The discovery of ferroelectricity in hafnium oxide (HfO2) in 2010 has sparked notable interest both in the scientific and industrial domains for ferroelectric memories. These devices offer significant advantages, including reduced energy consumption, good scalability to advanced technological nodes, high endurance, and fast read/write speeds.

This thesis aims to assess the electrical performance and reliability of non-volatile ferroelectric memory components based on HfO2 manufactured at CEA-Leti. The overall objective is to integrate these memories into advanced technological nodes to increase density while reducing energy consumption and operating voltages. Understanding the physical mechanisms responsible for the degradation of ferroelectric material properties will be crucial. To achieve this, various types of memory components ranging from single cells to matrices of several tens/hundreds of kilobits will be available for conducting this work. A portion of the thesis will also involve identifying obstacles to the integration of these materials into advanced technological nodes.