The ultimate impediment to the realization of any advanced image or data processing system is the limitation imposed
by inherent noise in the system. In this respect, nanostructured devices have a major advantage. Electronic noise in
nanodevices is qualitatively and quantitatively different from that in bulk devices. By exploiting both carrier and phonon
confinement effects, it is possible to suppress kinetic noise in quasi one dimensional structures by 2-3 orders of
magnitude in the microwave frequency regime. The excess noise is relocated to the infrared frequency regime where
electronic devices do not operate. This so-called "noise squeezing" is an aftermath of streaming oscillations in the
velocity and energy of current carriers, brought about by strongly peaked electron-phonon coupling at phonon emission
thresholds, and is a characteristic feature of one-dimensional carrier and phonon confinement.
Phonon confinement also dramatically alters the phonon dispersion relations and results in the preponderance
of phonon scattering events that involve high energy-low momentum phonons which can efficiently relax energy
without relaxing momentum. As a result, quantum 1/f noise is also strongly suppressed. The Hooge parameter,
which is a measure of the strength of 1/f noise in a system, is suppressed by at least one order of magnitude in
a quasi one-dimensional structures. This parameter (which is calculated within the framework of Handel's quantum
electrodynamic theory) also becomes relatively independent of the electric field driving transport.]A magnetic
field can further suppress 1/f noise by quenching backscattering events.
I will conclude this talk by showing that contrary to earlier speculation, quasi one-dimensional devices
(one dimensional MOSFETs, cylindrical surround gate transistors, etc.) do not typically exhibit any improvement
in mobility, transconductance or unity gain frequency. However, they can show dramatic improvement in noise
performance under almost any circumstance.
Work done in collaboration with Alexei Svizhenko, Mike Stroscio, Marc Cahay, Alexander Balandin and Kang Wang.
Continued interactions with Peter Handel are also gratefully acknowledged.
Copyright. All Rights Reserved.