Abstract:
After dominating the the semiconductor logic industry for decades the reign of Si is coming
to an end. Aggressive scaling of MOSFETs have pushed Si towards its physical performance
limits. By the end of this decade the Si MOSFETs will reach their theoretical limits. To continue
the scaling of logic devices and improve performance of integrated circuits researchers
must find a suitable replacement for Si. To meet the high current demands the replacement
must have high carrier mobility. III-V semiconductors appear as perfect candidates for
nMOS materials as they have very high electron mobility. But there are many possible III-V
materials. Also the aggressive scaling of Si has given birth to some very innovative structures
which solves very critical problems of nano-scale fabrication. Few of such innovative
structures are Double gate MOSFETs, Semiconductor On Insulator MOSFETs, FinFETs
etc. These structures might be utilized to achieve further performance enhancement of the
III-V based MOSFETs. Researchers all around the world are searching for the best possible
alternative using III-V materials. But fabrication of III-V materials pose some difficult
challenges. One such challenge is the gate dielectric. The III-V materials do not have aa
natural oxide dielectric like Si and thus growth of a dielectric with acceptable interface quality
is very difficult and perfecting one process is time and resource consuming. A theoretical
study of all the possible options of III-V MOSFETs will shed some light on how to choose
the most optimum path of device development and ensure perfect utilization of resources
and time. Thus theoretical evaluation of different performance markers of a MOSFET using
III-V material has become very important. There are several performance markers for a
MOSFET. The On current denotes the current driving capability of a MOSFET which in
turn decides the speed of a logic circuit. The subthreshold swing decides the power loss of a
MOSFET. The on/off current ratio also is a performance parameter of a MOSFET. As integrated
circuits are being shrunk down the speed of a transistor needs to increase while the
power consumption must decrease. Thus a systemic study of the On current of a MOSFET
will provide the engineers with one basis of selection or even lead to early elimination of some device structure that will not be able to sustain performance improvement in the long
run without wasting valuable time and resources. In this study a comparative picture of the
transistor On current for four different structures utilizing five different III-V semiconductor
is presented. The calculation has taken into account the quantum mechanical effects active
at the operating device dimensions. Only the limiting current is considered here and all
imperfections are ignored as the objective is to draw a comparative picture of the ultimate
performances of these devices.
