Magnetic Resonance in Nano-Scale Transistors

parkagendaΗλεκτρονική - Συσκευές

2 Νοε 2013 (πριν από 3 χρόνια και 9 μήνες)

63 εμφανίσεις

Magnetic Resonance in Nano
-
Scale Transistors


P.M. Lenahan

The Pennsylvania State University

University Park, PA 16801



The capabilities of integrated circuits have grown exponentially over the past four
decades, following Gordon Moore’s famous prediction
.

[1] This remarkable growth has
largely been the result of a continuous downscaling in individual device dimensions.
This downscaling has resulted in microprocessors with transistor gate dielectrics under
two nanometers; these small devices operate clos
e to some fundamental physical limits.
To cope with these limits and

sustain the Moore’s Law growth pattern, the materials
chemistry of small transistors has become more complex. The widespread introduction
of large quantities of nitrogen in gate silicon

dioxide and the replacement of silicon
dioxide by hafnium oxide (by Intel) are obvious examples of this growing
complexity
.
However
, the combination of complexity in materials and small device dimensions has
led to multiple

difficult

materials physics pr
oblems. Most of these problems involve deep
level defects which may capture charge carriers, take part in trap assisted tunneling, or
take part in poorly understood chemical reactions thought to be
triggered

by charge
capture. Since nearly all of these d
eep level defects may be rendered paramagnetic,
electron paramagnetic resonance (EPR) would seem an obvious technique of
choice

for
studies of these problems. Unfortunately, conventional EPR has a
sensitivity

of about
10
10

total defects.
(
Small transisto
rs don’t have
that many

defects.
)
Fortunately, several
electrically detected magnetic resonance (EDMR) techniques, spin

dependent
recombination (SDR) and spin dependent tunneling (SDT), offer sensitivities many
orders of magnitude higher than that of conve
ntional EPR. This presentation will deal
with the application of EPR mostly using EDMR, both SDR and SDT, to several
problems of current interest in the materials engineering of very small transistors
:

instabilities in small transistors with high nitrogen

content, and transistors with mostly
hafnium oxide gate stacks.


[1] Gordon E. Moore, “Cramming more Components onto Integrated Circuits
,”

Electronics
, Vol. 38, No. 8 (1965)