Date of Award

Spring 1-1-2013

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Conrad R. Stoldt

Second Advisor

Gordana Dukovic

Third Advisor

Yifu Ding

Fourth Advisor

Todd Murray

Fifth Advisor

Xiaobo Yin

Abstract

Given that iron pyrite (cubic FeS2, fool's gold) is a semiconductor with a ∼1 eV band gap, it has long been investigated for use in technological applications, especially photovoltaics. Unfortunately, numerous measurements indicate that it's properties, as currently synthesized at least, do not allow for effective devices. Photovoltages far below theoretical expectation are found as well as below band gap optical absorption. From a scientific standpoint, our understanding of the cause of these observations, the form of the density of states for instance, remains mired in uncertainty. In this work we have attempted to gain insight into this problem by creating ensembles of pyrite nanocrystals that can then be treated and measured with well-developed wet-chemical nanocrystal techniques. Specifically, we interpret the existing literature to advocate that the surface states of this material dominate its observed electrical properties. In an effort to better understand the most prevalent surface, the (100) face, we developed a synthesis that nucleates small (< 20 nm) pyrite nanoparticles and then changes chemical conditions to grow all other faces besides {100} to extinction, creating ∼37 nm nanocubes. The optical properties of these nanocubes are measured and the phenomenon of resonance light scattering (RLS) is observed. This phenomenon, along with the poor colloidal dispersibility of these nanocubes is then used to promote the idea that an unusual dynamic electronic phenomenon exists on these surfaces. This phenomenon is found to be passivated by introducing charged ligands to the surfaces of these particles. Additionally, after this surface treatment, two very sharp absorption features are observed at 0.73 and 0.88 eV. In connection with recent theoretical work, these transitions are taken as evidence that the (100) surface of pyrite is spin-polarized with each absorption peak being the signal of band edge absorption across a spin-selected direct band gap. A theoretical framework is proposed as a plausible explanation of the observed behavior. To wit, highly localized and energetically disordered Fe d-orbital states fill in the band gap of the (100) pyrite surface that is not perfectly terminated (or nearly so). Frustration between energetic disorder and Coulomb repulsion then results in the formation of metastable states that obscure the observation of these surface transitions and cause the dynamical behavior observed. It is further reasoned that one of these transitions, the one at 0.88 eV, has been observed before in cryogenic absorption and photoconductivity studies, and argued that a plausible reinterpretation of the data from these studies is possible. This reinterpretation can be rationalized within the context of the physical model posited here whereby cryogenic temperatures increase the importance of Coulombic interactions, which results in a decrease in the metastable DOS at the Fermi level and an electronic arrangement closer to that predicted theoretically, despite existing disorder. Finally, it is argued that the frustrated movement of electrons in metastable states can qualitatively explain the apparent conundrum in which surface sensitive probes are unable to measure the effect of isolated defects, despite the highly localized nature of the pyrite (100) surface. Given the range of phenomena this model explains, it may constitute a significant advance in our understanding of the electronic properties of pyrite. Additionally, given that electrochemical conversion of pyrite is a four electron process resulting in a high theoretical discharge capacity of 894 mAh g-1, we have synthesized micron-sized pyrite nanocubes for use in lithium-ion battery research. Previously the use of pyrite in such batteries has only been possible in non-rechargeable architectures. However, work described here shows that a solid state electrolyte can be used to contain the dissolution of these micron sized particles, allowing for battery cycling. This synthesis, especially the effect of pH on morphology, is described within the context of targeted requirements for a battery cathode material.

Share

COinS