Electrothermal Small-Signal Model of Nanosheet FETs With Zero-Temperature-Coefficient Based Parameters Extraction Method

Nanosheet FET (NSFET) is a promising structure for scaling transistors to the sub-5-nm node. However, the self-heating effect (SHE) impacts device performance at gigahertz frequencies, necessitating the small-signal modeling that accommodates SHE. Thus, an electrothermal coupled small-signal equivalent circuit model of NSFET and the corresponding parameters extraction method are introduced in this article. The introduced model is verified with TCAD simulated data, achieving excellent agreement between simulated and modeled
S
-parameters with a modeling error under 1.86%. A zero-temperature coefficient (ZTC)-based thermal network parameter extraction method is proposed, to accurately characterize SHE, enhancing stability of the extracted thermal resistance (
${R} _{\text {th}}$
) and capacitance (
${C} _{\text {th}}$
) at different
${V} _{\text {gs}}$
. The effectiveness of
${R} _{\text {th}}$
and
${C} _{\text {th}}$
is confirmed by small-signal model parameters extracted from TCAD data, achieving improved accuracy at low frequencies. The bias and temperature rise (
$\Delta $
T
) dependencies of intrinsic model parameters and underlying physical mechanisms are discussed. Results reveal that SHE negatively impacts
${g} _{m}$
and
${g} _{\text {ds}}$
, but positively affects
${C} _{\text {ds}}$
. Moreover,
$\tau _{m}$
is positively impacted at
${V} _{\text {gs}}$
below ZTC and negatively affected at
${V} _{\text {gs}}$
above ZTC. This introduced small-signal model provides valuable feedback for NSFET-based RF circuit design under SHE conditions.