| dc.description.abstract |
<p>Graphene, the wonder material, has captured the spotlight of electronics researchers
since its discovery. Even though it has proven to be unfit for use in digital transistors,
one of graphene’s lasting contributions is that it ushered in a whole new world of
two-dimensional (2D) materials. One sub-family of such 2D materials are transition
metal dichalcogenides (TMDs), which comprise ~40 different materials with a range
of electronic properties, including insulators, semiconductors, and conductors. Molybdenum
disulfide (MoS2) is one of the most studied semiconducting TMDs in recent years because
of its air stability, high effective mass and sizable bandgap ranging from 1.2 to
1.9 eV, which make it suitable for transistor applications. Challenges, however, still
cloud the promise of unbounded applications for MoS2 and other 2D TMDs. One major
difficulty is the formation of high-quality contacts as the performance of the 2D
transistors heavily relies on carrier injection through the contact interfaces. Although
some progress has been made in recent years, a robust, effective contact scheme is
still lacking and keeps the materials from being a viable option for future nanoelectronics.
</p><p>This dissertation presents systematic studies on the metal interface to 2D
materials for both the typical top-contact geometry and a newly developed in-situ
edge-contact geometry. First, in the top-contact geometry, two kinds of ion beam sources
(broad and convergent) are introduced to create defects in the otherwise dangling
bond-free surface of MoS2. Below a certain threshold, these generated defects are
shown to promote more efficient carrier transport between the contact metal and MoS2.
This ion beam modification approach decreases the contact resistance by 50% and doubles
the corresponding current in the device. Second, a pure edge contact scheme is introduced,
where an Ar ion beam is used in situ (with the metal deposition system) to etch the
MoS2, creating an abrupt edge profile that interfaces with the deposited contact metal
(without overlap of the metal on the MoS2). Edge contacts are able to withstand ultimate
scaling, down to sub-5 nm, and experimental results from edge contacts with a 20 nm
contact length (Lc) yield nominally the same performance as Lc = 60 nm edge contacts.
Hence, the in-situ edge contact approach shows great promise for aggressively scaled
transistor technology. </p><p>New observations on the contact scaling behavior of
top contacts is also covered. The transfer length of a typical metal-2D contact is
found to be much smaller than previously estimated by studying scaled contacts on
the same 2D crystal grain. Other progress in forming contacts to ultrasensitive 2D
materials is also presented, where the in-situ ion beam can expose air sensitive materials
for deposition of metal contacts without breaking vacuum. Based on these findings,
future work is proposed that includes 3D integration of 1D edge-contacted 2D FETs,
all-ALD transistors, and metallic CNT contacts to 2D materials. These future projects
will push the boundary of our understanding of scaled contact interfaces as well as
the development of emergent technologies for future nanoelectronic devices.</p>
|
|
