THE EXPECTED excellent performance of SiC devices is often expressed by a figure of
merit. In the past, several analyses of the impact of material parameters on the performance of
semiconductor devices have been performed. Johnson derived a figure of merit
(JFOM) [164]
(4.1)
which defines the power-frequency product for a low-voltage transistors. Here
is
the critical electric field for breakdown in the semiconductor and
is the
electron saturation velocity. Keyes' figure of merit (KFOM) [165] provides a thermal
limitation to the switching behavior of transistors used in integrated circuits
(4.2)
where is the thermal conductivity, is velocity of light and
is the static dielectric constant. These figures of merit predict that SiC is an excellent
material for high frequency devices. Baliga derived a figure of merit (BFOM) [166]
(4.3)
which defines material parameters to minimize the conduction loss in low-frequency unipolar
transistors. Here, is the mobility and
is the bandgap of the semiconductor. From this
figure of merit the excellent performance of high voltage unipolar devices in SiC can be
deduced. Baliga also derived a high-frequency figure of merit (BHFFOM) [167] for
unipolar switches
(4.4)
where
is the gate drive voltage and
is the breakdown
voltage. This figure of merit demonstrates that a significant power loss reduction can be achieved by
using SiC devices for high-frequency applications compared to other conventional semiconductor
devices.
Table 4.1:
Comparison of normalized figures of merit for -SiC and Si.
JFOM
KFOM
BFOM
BHFFOM
Si
1
1
1
1
4H-SiC
400
5.1
560
69
6H-SiC
400
5.1
240
29
Following these figures of merit (Table 4.1) and increasing
interest in high-temperature, high-power, and high-frequency devices based on SiC, the need for
numerical investigation pertaining to these devices becomes true. Device simulation has gained
increasing relevance for the design and optimization of electronic semiconductor applications
due to the rising design complexity and the cost reduction achieved by reducing the number of
experimental batch cycles. It has been a powerful and widely used tool in the investigation
and improvement of narrow bandgap semiconductor devices, and it will be a driving force in the
further development of both semi-conducting and semi-insulating SiC devices. A good survey on
the principle of device analysis and simulation is given in [104].
In
Chapter 3 material specific models and the corresponding
parameters relevant to -SiC material have been outlined and identified. It is the
scope of this chapter to evaluate the applicability of these models to state-of-the-art SiC
devices. The selection of new SiC devices for simulation studies are based on their
performance predictions which have been experimentally demonstrated in the past decade.
Subsections