In Situ Growth and Doping Studies of Topological Insulator Bismuth Selenide

Abstract

The past decade has borne witness to the rapid development of a new field

of theoretical and experimental condensed matter physics commonly referred to

as "topological insulators". In a (experimentalist's) single sentence a topological

insulator can be thought of as material that behaves like an empty metal box:

the i-dimensional material (i = 2,3) is insulating, but conducting states exist at

the (i-1)-dimensional boundaries. These edge or surface states possess non-trivial

properties that have generated interest and excitement for their potential utility in

solving practical problems in spin electronics as well as the creation of condensed

matter systems for realizing and testing new physics. The most extensively studied

materials systems with these properties suffer a major drawback in that the interior

of the metal box is not "empty" (insulating) but instead filled with metal. The

goal of this work has been to understand why the box is full instead of empty, and

explore a particular pathway to making it empty.

To address these outstanding questions in the literature pertaining to the persistent n-doping of topological insulator Bi2Se3, a custom apparatus was developed

for combined epitaxial thin film growth with simultaneous, in situ measurement

of transport characteristics (resistivity, Hall carrier density and mobility). Bi2Se3

films are found to be n-doped before exposure to ambient conditions and this doping

appears to be interfacial in origin. This work and methodology was extended to

study the efficacy of an amorphous MoO3 capping layer. MoO3 acts as an electron

acceptor, p-doping the Bi2Se3 up to a |∆n| = 1x10^13 cm−2

change in carrier density.

Complimentary X-ray photoemission spectroscopy (XPS) on bulk crystals showed

that this was enough to put the Fermi level into the topological regime. Thin films

were too highly n-doped initially to reach the topological regime, but the same magnitude change in doping was observed via the Hall effect. Finally, a Bi2Se3 film with

a 100 nm capping layer of MoO3 was vented to ambient, and found to retain a stable

doping for days of ambient exposure, demonstrating the effectiveness of MoO3 for

passivation.