.png){kind=icon}
.png){kind=icon}
Return
Nonlinear Asymmetric Parasitic Resistance Extraction of Vertical Pillar-Type FET Using Physics-Informed Artificial Neural Network
.png){kind=icon}
PRE
AI
J
S
S
D
D
J
S
Y
H
DOI：10.1109/led.2026.3701017.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In this study, a systematic decomposition of nonlinear asymmetric parasitic resistance (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {par}}\text {)}$ </tex-math></inline-formula> is presented for vertical pillar-type field-effect transistors, a promising architecture for achieving 4F<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>. By exploiting the forward mode and reverse mode resistance difference, which inherently cancels contact resistance, the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${R}_{\text {par}}$ </tex-math></inline-formula> is decomposed into three physically distinct components (bias independent-, drain field dependent-, and gate modulated-resistance). The extraction procedure relies solely on measured current-voltage data without requiring technology computer-aided design calibration (TCAD). To model the gate modulated component as a continuous function across the full bias space, a physics-informed artificial neural-network (PI-ANN) combining a separable parametric baseline with a neural-network correction is developed, ensuring extrapolation stability. The method is validated across pillar diameters from 0.40 to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$5.00~\mu $ </tex-math></inline-formula>m, revealing that the drain field-dependent component dominates across all diameters while the fixed component fraction increases with scaling. This framework provides a practical methodology applicable to compact modeling and design- technology co-optimization (DTCO).
Keywords:
Nonlinear parasitic resistance
physics informed-artificial neural network
vertical pillar-type FET
Journal
IF:
4.5
Papers:
614
Citations:
2.3W
Organization
Cited Papers