Effect of Environment on The Transport Properties of Graphene Devices

Open Access
- Author:
- Joshi, Prasoon
- Graduate Program:
- Electrical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 27, 2010
- Committee Members:
- Srinivas A Tadigadapa, Dissertation Advisor/Co-Advisor
Srinivas A Tadigadapa, Committee Chair/Co-Chair
Suman Datta, Committee Member
Theresa Stellwag Mayer, Committee Member
Jun Zhu, Committee Member - Keywords:
- Substrate Effects
Chemical Sensing
Nanotechnology
Graphene
Low - Dimensional Systems - Abstract:
- This thesis discusses the effect of environment on the transport properties of graphene devices. A process for making contacts to graphene, that is free of lithography protocols, has been devised. It has been shown that it is possible, using this process, to make two and four terminal graphene devices (chapter 2). In-vacuo investigation of the transport properties of graphene field effect devices show surface states of SiO2 can transfer electronic charge to graphene and it is expected that graphene devices on clean SiO2 surfaces be n-type. This is a view that is also supported by ab-initio molecular dynamic and density functional theory (DFT) calculations. Based on these experimental and theoretical results, a simple electrostatic model for graphene on SiO2 is presented (chapter 3). Gate voltage dependent hysteresis has been studied on graphene devices in which the SiO2 surface has been given different surface treatments that remove H2O molecules from the SiO2 surface. Two kinds of hysteresis are observed, charge neutrality point (CNP) hysteresis (due to loosely bound dipolar adsorbates) that gets suppressed upon dehydration, and transconductance hysteresis (due to near surface electron traps) that depends on the gate voltage sweep rate and is not suppressed upon dehydration of the devices (chapter 4). When n-type graphene devices are exposed to inert gases such as N2, Ar, He, Xe and Kr, the CNP is seen to shift to positive gate voltages, without much change in the minimum conductivity. The time scale of observation is of the order of days. Inert gases are not expected to dope the graphene devices. The CNP shift can be reversed by annealing the devices in vacuum. It is hypothesized that water molecules get trapped between the graphene and SiO2 and give rise to the CNP shift. A coverage of ~0.06 H2O molecules per hexagon is calculated based on the CNP shift (chapter 5). Graphene devices are exposed to dilute mixtures of NH3 in He. Upon exposure to NH3, the CNP of these devices are found to shift to negative gate voltages, indicating that NH3 is an electron donating molecule. Based on the CNP shift and themodynamic data for NH3 adsorption on graphite, a weak charge transfer of ~0.06e per NH3 molecule is calculated. Desorption of NH3 from the graphene device, as seen by recovery of the CNP, shows a double exponential behavior that resembles 1-D Fickian diffusion of NH3 at the graphene - SiO2 interface.