Ferroelectric (Hf, Zr)O2 Thin Films for High-Density Nonvolatile Memories
In the field of ferroelectric nonvolatile memories, the density enhancement approached the limit due to the scaling issues of perovskite-based ferroelectrics and complicated processing steps involved. However, the unexpected discovery of ferroelectricity
- PDF / 437,151 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 57 Downloads / 189 Views
Ferroelectric (Hf, Zr)O2 Thin Films for High-Density Nonvolatile Memories F. Ambriz-Vargas, R. Thomas and A. Ruediger
Abstract In the field of ferroelectric nonvolatile memories, the density enhancement approached the limit due to the scaling issues of perovskite-based ferroelectrics and complicated processing steps involved. However, the unexpected discovery of ferroelectricity in solid solutions of hafnium and zirconium binary oxides renewed the interest in high-density ferroelectric-based nonvolatile memories. This is mainly due to the familiarity of HfO2 and ZrO2 as a high-k dielectric material in advanced semiconductor devices and its CMOS compatibility. In this context, hafnium zirconate (HfZrO2) thin films were prepared on Pt/Al2O3/SiO2/Si substrates by radio frequency (rf) magnetron sputtering using a stoichiometric (Hf, Zr)O2 ceramic target. Fine-grained surface morphology for 500 °C deposited HfZrO2 films and rms roughness comparable to the underlying Pt surface were observed with atomic force microscopy (AFM). Films deposited at 500–750 °C showed typical hysteresis loop (phase angle vs. applied voltage) in the piezo force microscopy (PFM) studies, confirming the ferroelectric property of the film. Also, the stabilization of the ferroelectric phase at 500 °C is advantageous for the CMOS compatibility with TiN bottom/top electrodes.
⋅
Keywords Ferroelectric thin films (Hf, Zr)O2 Piezo force microscopy Nonvolatile memories
⋅
⋅
rf-magnetron sputtering
F. Ambriz-Vargas ⋅ R. Thomas (✉) ⋅ A. Ruediger (✉) Nanoelectronics-Nanophotonics, INRS-Énergie Matériaux et Télécommunications, 1650 Lionel-Boulet, Varennes, Québec J3X1S2, Canada e-mail: [email protected] A. Ruediger e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2018 M. Muruganant et al. (eds.), Frontiers in Materials Processing, Applications, Research and Technology, https://doi.org/10.1007/978-981-10-4819-7_12
123
124
F. Ambriz-Vargas et al.
1 Introduction Semiconductor electronics industry has been searching for a universal memory that is fast, nonvolatile, and high density for many decades. To realize such an ambitious task what is required: A memory (i) nonvolatile like FLASH, (ii) scalable to high density like as dynamic random access memory (DRAM) and FLASH, (iii) fast like static random access memory (SRAM), (iv) endurance (read-write cycles) like DRAM and SRAM. Along with that, the radiation resistant with low power consumption makes it even more exotic. However, these is no memory exist in the market which do the job of SRAM (cache memory), DRAM (main memory) and FLASH (storage memory). Classification of the various existing semiconductor memories along with the emerging memories are depicted in Fig. 1. Among the random access emerging memories, phase-change RAM (PRAM), magnetic RAM (MRAM), and resistance RAM (RRAM) are resistance-based memories, whereas ferroelectric RAM (FeRAM) is based on capacitance. Read-only memories (ROMs) are classified into two groups: (i) one-time programmable ROMs (OTPROMs) and (ii) erasable p
Data Loading...
