Defect Dynamics in Silicon-Doped HfO₂-Based Front-End-of-Line FeFETs: Insights From Low-Frequency Noise on Doping Concentration, Interfaces, and Write Cycling

This article comprehensively examines the defect distribution and dynamics through low-frequency noise (LFN) characteristics in hafnium-oxide (HfO<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub>)-based ferroelectric field-effect transistors (FeFETs). We delve into the intricate effects of varying silicon (Si) doping levels of the hafnium (<roman xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hf</roman>)-based ferroelectric, the choice of interfacial layer (IL) materials, and the implications of write cycling on LFN performance. Our findings indicate that increased Si doping levels significantly reduce carrier mobility fluctuations while preserving stable and consistent defect distributions. Notably, devices incorporating silicon-oxynitride (<roman xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SiON</roman>) as an IL between Si channel and HfO<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> improve the noise immunity of FeFETs compared with those with Si dioxide as an IL, a benefit attributed to enhance defect passivation properties present in <roman xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SiON</roman>. In addition, the advantages and dependencies of different doping concentrations, as well as two different interface materials related to the noise behavior, are discussed with regard to the physical view of the flicker noise investigation method based on the model of Ghibaudo. Moreover, the endurance testing reveals a crucial shift in defect distributions, transitioning from predominantly bulk traps to a predominance of interface traps. This shift directly impacts the memory window’s (MW) closure, potentially influencing device performance over prolonged use. This study ultimately contributes valuable insights toward optimizing FeFET technology, particularly in the context of advanced memory applications, paving the way for more reliable and efficient memory storage solutions.