Three-Dimensional Electrostatics- and Atomistic Doping-Induced Variability of RTN Time Constants in Nanoscale MOS Devices-Part II: Spectroscopic Implications

• Published in: IEEE transactions on electron devices Vol. 59; no. 9; pp. 2495 - 2500
• Main Authors: Monzio Compagnoni, Christian, Castellani, Niccolò, Mauri, Aurelio, Spinelli, Alessandro S., Lacaita, Andrea L.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01-09-2012; Institute of Electrical and Electronics Engineers
• Subjects: Applied sciences; Atomistic doping; Compound structure devices; Correlation; Doping; Electron traps; Electronics; Electrostatics; Exact sciences and technology; Flash memory devices; Integrated circuits; Integrated circuits by function (including memories and processors); Logic gates; Magnetic and optical mass memories; Molecular electronics, nanoelectronics; Nanoscale devices; random telegraph noise (RTN); semiconductor device modeling; Semiconductor electronics. Microelectronics. Optoelectronics. Solid state devices; Storage and reproduction of information; Substrates; variability; Memory devices; Positioning; Atomistic doping; Doping; MOS structure; random telegraph noise (RTN); Information extraction; variability; Random telegraph noise; Modeling; Nanoelectronics; Flash memory devices; Semiconductor device; Three dimensional model; Non volatile memory; Flash memory; semiconductor device modeling; Integrated circuit; Time analysis; Time constant
• ISSN: 0018-9383; 1557-9646
• DOI: 10.1109/TED.2012.2203412

This paper investigates the impact of 3-D electrostatics and atomistic doping on the spectroscopic analysis of random telegraph noise (RTN) traps in nanoscale MOS devices. Using the numerical model and the template decananometer Flash cell presented in Part I of this paper, the gate bias dependence of the capture and emission time constants of oxide traps is shown to largely depend on the trap position over the channel, both in the subthreshold and in the on-state regime. This compromises the accuracy of any 1-D method for trap spectroscopy based on the time constants analysis and, due to the randomness in trap position and dopant placement in the substrate, calls into question the possibility for any accurate trap spectroscopy in nanoscale devices. Finally, the possibility to extract any information on the trap depth from the fluctuation amplitude of RTN waveforms is shown to be precluded by the large statistical spread of the amplitude itself, resulting in its negligible correlation with the trap position in the oxide and with the waveform time constants.