Thesis: Ph. D., Massachusetts Institute of Technology, Department of Materials Science and Engineering, 2013. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 105-114). / Despite its exceptional optical absorptivity, suitable band gap, and earth abundance, the low open-circuit voltage of pyrite FeS₂ has remained the biggest challenge preventing its use in photovoltaic devices. Two widely-accepted causes are: (i) Fermi level pinning caused by intrinsic surface states that appear as gap states; (ii) presence of the polymorph marcasite. Based on density-functional theory (DFT) calculations, (i) the intrinsic (100) surface states are not gap states but located at the conduction band edge; (ii) epitaxial growth of marcasite on pyrite is thermodynamically favorable, but its band gap (from Kohn-Sham [delta]-sol method) is not less than pyrite. It is unlikely that the photovoltaic performance of pyrite is undermined by intrinsic surface states or marcasite. The stoichiometry and the ubiquitous observation of unintentional p-type conductivity of pyrite thin films are investigated via DFT defect computations. Native defects occur in low concentrations due to high formation energies, implying that pyrite is intrinsically stoichiometric. The p-type conductivity can be caused by OS defects under oxidizing conditions. Band gap engineering of pyrite is studied by alloying with non-rare-earth isovalent elements via DFT computations. We identify six MS₂ candidates that have larger band gaps than pyrite. Band gap enhancement of pyrite is observed only in the Ru and Os alloyed systems, but their incorporation into pyrite may be severely limited. All other candidate alloys exhibit large gap bowing effects due to size and/or electronegativity mismatch. The effects of particle size and pH on the relative phase stability of pyrite and marcasite polymorphs are explored. The size effect is incorporated through volume scaling of Wulff shapes. The pH effect is modeled by generalized, charged surface energies as a result of ion adsorption from the aqueous environment. Based on joint density-functional theory calculations, pyrite is unstable in highly acidic conditions due to a negative H+-adsorbed (110) surface energy, but stabilized for pH >/~ 2. Directions for future work are briefly discussed. / by Ruoshi Sun. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2008. / Includes bibliographical references (p. 157-162). / The solid oxide fuel cell (SOFC) with its high energy conversion efficiency, low emissions, silent operation and its ability to utilize commercial fuels has the potential to create a large impact on the energy landscape. Although SOFCs do not require noble metal catalysts, it has the disadvantage of having to operate at elevated temperatures (>800°C) due to the poor catalytic activity of the oxygen reducing cathode. This thesis research is focused on identifying the fundamental reaction mechanisms and pathways for oxygen reduction reaction (ORR) in transition metal oxide cathodes to enable lower operating temperature and improve device performance. The approach used here involves dense and thin-film microelectrodes of Lao.sSr0.2M03-d (M = Mn and Fe) supported on thin film 8YSZ electrolyte that was fabricated via microfabrication/photolithography techniques. These microelectrodes were then probed mainly via electrochemical impedance spectroscopy (EIS) under varying temperature, oxygen partial pressure, microelectrode size/thickness, and applied polarization potential to determine the rate-limiting step(s) during ORR. For La0.8Sr0.2MnO3-d (LSM), at least four reaction or transport process have been identified from EIS; (i) ion conduction in 8YSZ, (ii) possible surface diffusion on LSM, (iii) surface exchange reaction on LSM and (iv) TPB/bulk charge transfer on LSM. The rate-limiting step and the relative contribution of ORR current from the bulk and three-phase boundary pathways were found to vary with temperature. This was quantified by correlating microelectrode geometry to the impedance. For Lao.sSr0.2FeO3-d (LSF), overall impedance was found to be at least one to two of magnitude smaller than LSM. / (cont.) Three reactions or transport processes were observed that were attributed to (i) ion conduction in 8YSZ, (ii) possible surface diffusion on LSF and (iii) oxygen surface exchange reaction on LSF. The bulk pathway was found to be dominating down to 570°C and surface exchange as the rate-limiting step. Studies to probe the surface reaction mechanisms showed that oxygen exchange coefficients (kchem) were highly dependent on material electronic properties rather than oxygen vacancy contents. Using polarization potentials, kchem was found to be increased by up to 1 order of magnitude due variations in lanthanum elemental contents on the microelectrode surface. Design principles for thin-film based cathodes are proposed to enhance performance of SOFCs operating at <700°C. For dense and thin-film electrodes, enhancement of surface exchange rates via physical or chemical modifications is critical to enhancing surface catalytic activity. Use of internal microstructure with large grain boundary fractions to improve bulk oxygen transport is important, especially in LSM, for reducing ion transport losses. / by Gerardo Jose Cordova la O'. / Ph.D.

Thesis (M. Eng.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2009. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student submitted PDF version of thesis. / Includes bibliographical references (p. 142-150). / In this thesis report, both quantitative and qualitative approaches are used to provide a comprehensive analysis of lithium ion (Li-ion) batteries for plug-in hybrid electric vehicle (PHEV) and battery electric vehicle (BEV) from technological and economical perspectives. Five key factors including power density, energy density, safety, durability, and cost are employed to compare four types of Li-ion batteries. Utility analysis indicates that all the Li-ion batteries are able to satisfy both power density and energy density targets, but only two of them are able to meet safety and durability requirements. Currently, the main challenge for their automotive application is cost reduction, since the cheapest LiFePO₄ battery costs $247.8/kWh which is 1.65 times the cost target established by USABC. Economical values of PHEV and BEV are presented from an end user's point of view. Various sensitivity analysis have been used to identify the impact of key factors such as battery pack cost reduction, driving distance, gasoline price, and government subsidizations on cost effectiveness of PHEV and BEV. Results show that $4,270 and $7,726 of U.S. government subsidizations to an individual user are needed for PHEV and BEV to breakeven. Lastly, the lithium ion battery based electric vehicle systems have also been evaluated in the implementation models in Singapore. The conclusion is that it is not feasible to adopt electric vehicle system in Singapore under current government incentives. / by Haitao Fu. / M.Eng.

Thesis (S.B.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2007. / Includes bibliographical references (leaves 41-43). / Pulsed laser deposition was used to create thin films of Ce-Fe-O and Y-Fe-O systems. Deposition temperature and ambient oxygen pressure were varied systematically between samples to determine which deposition conditions were most favorable to the formation of cerium/yttrium orthoferrite. The structure and composition of each film were then determined using X-ray diffraction and wavelength dispersive spectroscopy respectively. In addition, the magnetic and optical properties of the yttrium films were characterized to determine the suitability of these materials as Faraday isolators at A=1550 nm. Results show that orthoferrite crystal structures in these systems are not stable in the temperature and oxygen ranges tested. It was also found that increasing oxygen pressure caused exponential decay in the deposition rate. Most films were amorphous, exhibiting a paramagnetic M-H plot and a Verdet coefficient between 0.37 and 0.89 deg cm-1 Gauss-1 / by William Wagner Supplee, Jr. / S.B.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2010. / Vita. / Includes bibliographical references. / Biological building blocks served as excellent templates for the preparation of various nano-materials due to their beneficial interactions at the molecular level. The bio-mineralization of genetically engineered M13 bacteriophage resulted in one-dimensional nanowires having outstanding properties in diverse applications. As a bridge between the chemical synthesis of nanostructures and the bio-mineralization of M13 phage, surfactant molecules were introduced to the biological systems. The specific affinity of M13 phage with Au-binding peptides was strong enough to attract Au ions despite the existence of surfactant molecules. Consequently, the surfactant-mediated bio-mineralization of M13 phage enabled us to precisely control the morphologies and structures in nanometer scale. The Au-binding M13 phage could also integrate other noble metals (Ag/Pt/Pd) to prepare homogeneous Au-based noble metal alloy nanowires in structures and compositions, and their electrochemical properties upon the systematic changes in compositions were investigated. Especially for the Au-Pt system, the catalytic activity study on the two distinct structures, the alloy and the core/shell, provided us important factors to design new catalysts with optimized activities. / by Youjin Lee. / Ph.D.

Thesis: S.M. in Materials Science, Massachusetts Institute of Technology, Department of Materials Science and Engineering, 2014. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 90-92). / Active hyperspectral imaging (HSI) is a promising technique for the detection of chemicals at standoff distances. In active HSI, a target is illuminated by a laser source at many different wavelengths and a camera obtains an image of the illuminated scene at each wavelength. In this research, the feasibility of hyperspectral imaging for the detection of particles on surfaces was demonstrated using potassium chlorate particles on car panels at distances of 5 m, 10 m, and 20 m. Using the Adaptive Cosine Estimation (ACE) algorithm which compares the observed reflectance spectra to a reference spectrum, potassium chlorate fingerprints are easily visible at many different sample angles. However, in general, there is a large amount of variation in the shape and magnitude of spectra in a hyperspectral image that depend on factors such as particle size, viewing geometry, and surface reflectivity. Thus, Mie Theory calculations are performed on simulated materials and combined with information from sources such as Hapke [4], [20] to give qualitative insight into the expected shape and magnitude of reflectance spectra from sparse particles on a surface. The shape of the spectra depends on whether the particles are strongly absorbing or weakly absorbing. Strongly absorbing particles tend to have reflectance maxima near the resonant frequency, whereas weakly absorbing particles tend to have reflectance minima. For highly reflective substrates, the reflectance decreases sharply as the sample angle increases and becomes dominated by backward scattering from the particle which has a flatter spectrum around the Christiansen frequency, the frequency at which the refractive index of the particle is closest to one. The double interaction model, which uses Mie Theory to calculate the contributions to the reflectance along two different light paths, is used to accurately account for how the shape and magnitude of the reflectance spectra of sodium chlorate particles on gold and silica surfaces changes as a function of sample angle and laser angle. A method for approximating the mean particle size based on the location of the peak near the Christiansen frequency is derived. This method, when applied to the sodium chlorate sample, yields a result for the mean particle diameter that is approximately half of the value determined using a microscope. The Hapke Isotropic Multiple Scattering Approximation (IMSA), combined with Mie Theory, is used to give qualitative insight into the expected shape and magnitude of reflectance spectra from bulk powders. Compared with the reflectance spectra from sparse particles, the spectra from bulk powders are much simpler and less dependent on the viewing geometry. The Hapke IMSA model is able to accurately account for the observed changes in the reflectance from bulk sodium chlorate powder at multiple sample angles and laser angles. A final scenario of interest is thin films on rough or porous surfaces. Using a model that takes into account diffusely reflected and specularly reflected light, the observed reflectance spectra from diethyl phthalate (DEP) on a brick is fitted to a high degree of accuracy. This suggests a promising method for using hyperspectral imaging to determine the thickness of liquids on porous surfaces. Finally, the issue of speckle in hyperspectral imaging was examined using simulations based on Fourier optics and information from sources such as Goodman [6], [17]. Speckle is a limiting factor in hyperspectral imaging because it is noise that scales with the signal, and thus cannot be eliminated by increasing the signal strength. Equations from various sources are presented that describe the reduction in speckle contrast for spatial, spectral, polarization, temporal, and angular averaging. Original equations for the reduction in contrast for spectral and angular averaging are derived. / by Travis R. Myers. / S.M. in Materials Science

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2002. / Includes bibliographical references (leaves 52-54). / The unprecedented growth of the semiconductor industry is demanding ultra large-scale integrated (ULSI) circuits with increasing performance at minimum cost and power dissipation. As the critical dimensions in ULSI design continue to shrink, system performance of integrated circuits will be increasingly dominated by interconnect delay. Three-dimensional (3-D) ICs can reduce interconnect delay problems by offering flexibility in system design, placement and routing. 3-D ICs can be formed by vertical integration of multiple device layers using wafer bonding, recrystallization or selective epitaxial growth. The flexibility to place devices along the vertical dimension allows higher device density and reduced total interconnect lengths in 3-D ICs. One approach to fabrication of 3D integrated circuits is to bond previously-processed device layers using metal-metal bonds that also serve as layer-to-layer interconnects. Evaluation of the feasibility of wafer bonding for 3-D integration relies on our ability to characterize bonded interconnects. The reliability of devices containing multi-layer thin film structures is strongly influenced by the adhesion properties of the many interfaces present. Interface fracture failure is highly likely given the high thermal stresses developed during processing and also during service. A four-point bend test technique has been used to evaluate the strength of Cu-Cu bonds. Test structures were fabricated by bonding wafers containing copper lines (with Ta barrier) that were patterned on silicon dioxide. Tests on the thermocompression-bonded copper lines yielded reproducible fracture toughness values (1-10 J/m2 ) for bonds created at 300°C-400°C. The effect of process parameters on bond strength was studied. It was found that surface copper oxide removal prior to bonding using a forming gas purge (95%Ar-5%H2 ) resulted in higher toughness values and lesser variations compared to a N2 purge. Also, bond strength was found to increase with increasing bonding temperature. Thicker bonded films resulted in stronger bonds. Interface failure was found to be most likely at the Cu-Cu and Ta-Silicon dioxide interfaces. The results obtained from different process conditions were used to optimize the bonding process. / by Rajappa Tadepalli. / S.M.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2004. / Includes bibliographical references (leaves 108-114). / (cont.) given slurry was proposed. The fluidity of rheocast A357 alloy slurries was contrasted with the fluidity of superheated liquid. Rheocast slurries with 37% solid particles,were found to flow about half as far as fully liquid alloy superheated 20⁰C above the liquidus. / The importance of semi-solid metal processing derives primarily from its ability to form high integrity parts from lightweight alloys. Since the discovery of the semi-solid metal microstructure, most part production was by reheating of billets which possessed a suitable microstructure ("thixocasting"). However, it is now apparent that there are significant advantages of forming semi-solid slurry directly from liquid alloy ("rheocasting") and efficient rheocasting processes have been engineered. In this work, experimental and analytical approaches were taken to study how non-dendritic microstructures form and evolve in Al-4.5wt%Cu alloy during the earliest stages of solidification. Experimental results showed that particles in quenched rheocast alloy were already spheroidal, and free of entrapped eutectic, after 5 seconds of solidification time. Spheroidal particles were also formed by reheating equiaxed dendrites of approximately 10 [micro]m radius above the eutectic temperature for 5 seconds, but these spheroids contained entrapped eutectic. In both rheocasting and reheating experiments, the average particle radius was found to increase with solidification time at a rate that closely follows the classical dendrite arm ripening curve. Particle growth models developed were compared with the average particle radius measurements, and particle solute content measurements. The maximum cooling rate to maintain spheroidal interface stability at various solid fractions was studied experimentally. A modified stability model which considered particle interaction through solute field overlap was developed and found to be in good agreement with experimental data. A simple method for the foundry to determine the maximum cooling rate for a / by Raul A. Martinez-Ayers. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1995. / Vita. / Includes bibliographical references (leaves 217-218). / by Hongbao Liu. / Ph.D.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1998. / Vita. / Includes bibliographical references (leaves 112-115). / by Prashant Soral. / S.M.