< Back to previous page

Project

Understanding Charge Behaviour in a 2D Transition Metal Dichalcogenide MOS System

Advancements in technology are driven by downscaling the channel length and
the thickness of semiconductor that improve performance of a MOSFET. 2D
semiconductors, like transition metal dichalcogenides (MX2), are van der Waals
(vdW) layered structures with one layer approx. 0.7 nm thick and self-terminated
surfaces with no dangling bonds. They offer the promise of ultrathin channels
with high mobility for future technology nodes. 2D materials have a wide range
of physical and chemical properties, which has also generated keen interest in
the sensors and display technology communities. However, in reality, material
quality, processing issues and non-ohmic contacts have resulted in sub-standard
performance of 2D MOSFETs. It is compounded by an inaccurate understanding
of electrostatics and charge transport in the channel.

In this thesis, we investigate a fundamental aspect of any 2D transistor −
the oxide-semiconductor interface. Using MOS admittance spectroscopy, we
inspect interface and oxide defects (Dit and Nbt, respectively) in MoS2 flakes
and MOCVD grown MoS2 and WS2. We propose a new electrical test
structure, called the edge MOS capacitor, which can probe defect states in the
MX2 bandgap. Due to its design, MOS admittance characteristics become very
sensitive to the channel length of the edge MOS capacitor (Lch) and the in-plane
channel resistance (Rch). Using the Silicon MOS capacitor model results in an
overestimation of Dit as it does not factor the Lch and Rch dependence.

Therefore, we develop a new MX2 MOS capacitor model with 2D electrostatics
in DC and a distributed network in the AC regime. Using this, we study the
strong coupling between Rch and vertical gate electrostatics in an edge MOS
capacitor structure. We find that true Dit can be extracted only from short
channel devices, while long channel devices are useful to evaluate electron/hole
mobility. Using this an exponentially decaying Dit from the conduction band
edge is found at the MoS2-HfO2 and MoS2-SrTiO3 interface.

We also find that the maximum charge density (in accumulation) is limited by interfacial residues at the oxide-semiconductor interface. To boost channel
performance, we investigate top and dual gate devices using MOCVD MoS2 and
WS2. The true potential of a dual gate device can be realised in a scaled top
and bottom gate configuration. We also observed hole transport in dual gate
WS2 grown on SiO2 with edge contacts, paving way for ambipolar an p-type
FETs using 2D materials.

Date:1 Feb 2016 →  7 Jul 2020
Keywords:2-D materials, TMD, MoS2
Disciplines:Nanotechnology, Design theories and methods
Project type:PhD project