Detailseite
Projekt Druckansicht

Kompaktmodellierung und Charakterisierung von Kohlenstoff-Nanoröhren-Transistoren sowie Entwurf analoger Hochfrequenzschaltungen

Antragsteller Professor Dr.-Ing. Michael Schröter, seit 10/2017
Fachliche Zuordnung Elektronische Halbleiter, Bauelemente und Schaltungen, Integrierte Systeme, Sensorik, Theoretische Elektrotechnik
Förderung Förderung von 2013 bis 2020
Projektkennung Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 232931620
 
Erstellungsjahr 2020

Zusammenfassung der Projektergebnisse

A closed-form analytical drain current formulation based on the Landauer integral was developed that takes into account electron–phonon scattering in the channel and was successfully used for performance predictions from varying Schottky barrier (SB) height, characteristic length, electron mean free path and band gap of the channel material. The new formulation removes deficiencies in the treatment of the effective mean free path of electrons scattered by acoustic phonons in previously existing compact models. Unfortunately, project funding was insufficient for developing a complete CNTFET model with an accurate description of the model parameters as a function of CNT density and diameter as well as device geometry and substrate coupling and thermal effects. It was also shown that the generic formulations of a "virtual source" compact model, which has become popular for nanoscale silicon MOSFETs, is quite reasonably suited for capturing the measured DC and high-frequency (HF) characteristics of CNTFETs with different channel lengths as well as of 200 nm III-V (InGaAS) nanowire HFETs and 120 nm Si based MOSFETs. However, the physics of the source and drain SB are not included in this model, so it lacks predictive capability and is also not capable of handling the characteristics of high-linearity CNTFETs. Based on the developed compact models, which were calibrated on measured data, CNTFET- based high-frequency amplifiers were studied regarding the impact of transistor linearity and CNT inductor quality factor Q on maximum gain and low-noise applications at 2.4 GHz. The results of the designed single- and double-stage amplifiers indicate that the performance is not affected as long as Q is beyond 8. Hence CNT-based inductors appear to be feasible components for RF amplifiers. Comparing the noise and gain performance of these designs to other FET technologies and considering the linearity performance, CNTFET-based amplifiers turn out to be credible contenders for developing circuit applications at S-band frequencies using on-wafer stabilizing and matching networks. In addition, a CNTFET based oscillator operating at 461 MHz (−115 dBc/Hz phase noise at 1 MHz) was designed, implemented and characterized for the first time. The thermal resistance and capacitance of fabricated HF CNTFETs were determined for the first time. The extracted parameters, especially the short thermal time constant in the range of a few ps, confirm the thermal stability expected for these devices. The developed lumped element electro-thermal model agrees well with the experimental data obtained from pulsed measurements. The original plan in this project was to apply the models to a larger variety of HF CNTFETs with also significantly improved performance. Unfortunately, funding of CNTFET process development has been inadequate for implementing and running a stable wafer-scale process flow for test chip fabrication. This has so far limited the experimental data. Different channel materials were compared with the aim of identifying the potential of a CNT based FET technology. The results show that with improved contact resistance and tube density, CNTFETs can reach far higher cut-off frequencies than FETs with incumbent channel materials, including III-V channels. Along with their much higher transconductance and simple fabrication process, CNTFETs would then lead to superior circuit performance and hence become the FET technology of choice.

Projektbezogene Publikationen (Auswahl)

 
 

Zusatzinformationen

Textvergrößerung und Kontrastanpassung