Small-signal parameters extraction and noise analysis of CNTFETs

TL;DR

This paper presents a technique for extracting small-signal parameters and modeling noise in CNTFETs, enabling accurate high-frequency performance prediction for microwave circuit applications.

Contribution

It introduces a comprehensive equivalent circuit model and noise analysis method specifically tailored for CNTFETs, validated against experimental data.

Findings

Good agreement between model and experimental data

Shot noise dominates total noise due to Schottky barriers

Model effectively predicts high-frequency performance

Abstract

The use of carbon nanotube (CNT) field-effect transistors (FETs) in microwave circuit design requires an appropriate, immediate and efficient description of their performance. This work describes a technique to extract the parameters of an electrical equivalent circuit for CNTFETs. The equivalent circuit is used to model the dynamic and noise performance at low- and high-frequency of different CNTFET technologies, considering extrinsic and intrinsic device parameters as well as the contact resistance. The estimation of the contact resistance at the metal/CNTs interfaces is obtained from a Y-function based extraction method. The noise model includes four noise sources: thermal noise, thermal channel noise, shot channel noise and flicker noise. The proposed model is compared with a compact model calibrated to hysteresis-free experimental data from a high-frequency multi-tube (MT)-CNTFET technology. Additionally, it has been applied to experimental data from another fabricated MT-CNTFET technology. The comparison in both cases shows a good agreement between reference data (simulation and experimental) and results from the proposed model. Low- and high-frequency noise projections of the fabricated reference device are further studied. Noise results from both studied technologies show that shot noise mainly contributes to the total noise due to the presence of Schottky barriers at contacts and along the channel.