Unleashing the Potential of Integrated Ferroelectric Devices with Hafnium Oxide

Abstract

Early ferroelectric materials discovered more than 100 years ago like Rochelt salt were too unstable for widespread applications. Perovskite ferroelectrics like BaTiO3 (BTO) and later PbZrxTi1-x O3 (PZT) opened the path for widespread applications. Fist memory devices based on cross-point arrays involving BaTiO3 were proposed in the early 1950s [1]. However, it was not until semiconductor technology became mature that the first integrated capacitor-based memories involving PZT ferroelectrics became available in 1993 [2]. Ferroelectric memories remained a niche market due to the complexity of integrating oxide perovskites into CMOS and the related slow scaling. In 2011 Boescke et al. [3] published the observation of ferro electricity in hafnium oxide. This new material system has strongly increased the possibility to integrate ferroelectric functionalities into CMOS processes in the following years. Fig. 1 illustrates this development by showing the publication activity on the topic of “ferroelectric memory”.