Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2004. / Includes bibliographical references (p. 199-205). / In this work, we examine fundamental materials processes in the growth of indium gallium phosphide (InGaP) via metalorganic chemical vapor deposition (MOCVD). In particular, we realize improvements in the epitaxial integration of high-quality InGaP device materials on non-standard platforms, such as GeSi graded buffers orSi substrates, and InGaP or indium aluminum gallium phosphide (InAlGaP) graded buffers on GaP substrates. We apply these improvements to the design and implementation of strained-InGaP quantum-well light emitting diodes (LEDs) operating in the yellow-green region of the visible spectrum. The innovative use of these traditional materials is intended to provide a solution for bright green solid-state light emitters. Initial modes of InGaP lattice-matched epitaxy on GeSi were studied. Three- dimensional growth was observed over a wide range of deposition temperatures and V/III ratios. Pre-growth thermal cycling in a H2 plus PH3 ambient led to a large increase in surface roughness and the formation of surface mesas. Thermodynamic simulations suggest that these mesas may be P clusters or GeP solid complexes. They may also be surface oxides formed in conjunction with water vapor in the deposition chamber. Such surface degradation prior to the initiation of epitaxy is unfavorable for monolayer growth. The development and evolution of defect microstructures in relaxed, compositionally graded InGaP buffers deposited on GaP were examined. In particular, the properties of branch defects in InGaP graded buffers were examined for a large number of growth and annealing conditions. / (cont.) These studies confirm that branch defect formation is driven by surface, not bulk, processes. Branch defects in the bulk arise from surface features that are metastably "frozen" in place by subsequent deposition and propagate through the thickness of the sample. We conclude that branch defects comprise a local compositional fluctuation resulting from the clustering of In atoms. This identification is supported by the suppression of branch defect formation under conditions of reduced adatom mobility, including low growth temperature and high V/III ratio. In addition, we demonstrate that dislocations gliding in the [110] direction are -preferentially blocked by strain fields arising from nearly-[110]-oriented branch defects. This is further evidence for the link between branch defects and In clustering. A relaxed InAlGaP graded buffer platform was utilized in the design and fabrication of a novel strained-InGaP quantum-well epitaxial-transparent-substrate LED (ETS-LED). The best devices exhibited yellow-green emission with a primary wavelength of 590 nm and a secondary wavelength of 560 nm. These devices had [rho]TD = 7 x 106 cm2 and an In 0.32 Ga 0.68 P quantum well active region, and operated at 0.18 [mu]W per facet at 20 mA, corresponding to a luminous efficacy of approximately 0.01 1m/W. Transmission electron diffraction indicates that the observed spectral lineshape is the result of emission from ordered and disordered domains in the quantum well. Devices with [rho]TD = 5 x 107 cm-2 and an In 0.32 Ga 0.68P ... / by Lisa Megan McGill. / Ph.D.

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1996. / Includes bibliographical references (p. 101-104). / by Ching-Te Lin. / M.S.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Materials Science and Engineering, 2016. / 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 133-150). / Alloys based on face-centered cubic (FCC) elements such as nickel (Ni) are among the most resistant to fracture. However, when embrittled by impurities, they lose their toughness and crack along grain boundaries. Though long known, this phenomenon remains poorly understood. In this thesis, we use large-scale molecular dynamics (MD) simulations to study the effects of grain boundaries (GBs) on various aspects of fracture properties in Ni, including intergranular fracture mechanisms, fracture toughness as well as crack healing. By performing statistical analysis on crack tip processes for fracture along different GBs, we revealed three distinct crack propagation mechanisms. For fracture along coherent twin boundary with the crack front along the [112] direction, no bond breaking is observed and crack advance is solely attributed to the slip of atoms at its tip due to the emission of dislocations. The dislocation process leads to the blunting of the crack tip. For fracture along [Sigma]265(100) symmetrical tilt GB, we discovered a new crack propagation mechanism, decohesion restrained by emission of dislocations (DRED). In it, bursts of brittle fracture initiate emission of dislocations, which pre- vent cracks from propagating more than a few nanometers in a single burst. For fracture along coherent twin boundary with the crack front along the [110] direction, crack propagates by brittle decohesion, which initiates dislocation emission in a similar way as DRED. However, the dislocation process does not arrest the crack due to the local hardening mechanism, which constraints the motion of dislocations. Using the method developed to calculate the critical energy release rate Gc from atomistic simulations, we also compared the toughness of fractures by these three mechanisms. In the course of investigating intergranular fracture, we discovered a new mechanism for crack healing in a 2D model. This mechanism relies on the generation of disclination dipoles due to GB migration, which can interact with the crack, causing it to advance or heal. We also demonstrate the healing of nanocracks in realistic 3D microstructures. / by Guoqiang Xu. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1995. / Includes bibliographical references (p. 211-217). / by Orhun Kamil Muratoğlu. / Ph.D.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2001. / Includes bibliographical references (p. 99-102). / We present an extensive First Principles study of structure, defects and proton insertion in Mn02. It is shown that the paramagnetic extrapolations of spin-polarized results are essential to correctly reproduce pyrolusite as the ground state of Mn02. While many other structures are found to be near degenerate in energy with pyrolusite, no thermal disorder exists in the system up to several thousand degrees as the strong correlation of the Mn-vacancy order along the lines of face sharing octahedra removes any low-energy excitations from the system. Mn-vacancies compensated by protons, ubiquitously present in commercial Mn02 have a dramatic effect on phase stability and induce the formation of ramsdellite Mn02 and twinning defects. We believe these proton compensated Mn vacancies to be the source of the structural complexity of synthetic Mn02 produced either electrochemically or chemically. It is shown that protons are always covalently bonded to an oxygen atom in Mn02. In ramsdeHite, the proton prefers the pyramidal oxygen to the planar coordinated oxygen atom. In both pyrolusite and manganite, the protons may appear to be at an octahedral center in experiments as the activation barrier for hopping between the two stable sites on each side of the octahedral position is only about 25 meV. Introduction of die Wolff disorder and twinning defects is found to have a large adverse effect on the diffusivity of protons in [gamma]-Mn02. Protonation also increase barriers to proton migration due to Jahn-Teller distortion and H-H interactions. Results indicate that direct H-H interactions are not that significant compared to oxygen mediated indirect interactions, observed in manganite. Experimental and calculated ramsdellite discharge curves deviate significantly at the early stages of the reduction process. We believe that a significant source of this discrepancy is the presence of proton compensated Mn vacancies in real Mn02, which create local sites with higher discharge potential. Calculations also suggest that the ordered phase, observed in experiments at mid-reduction (groutellite, MnOOHo.5), may be due to lattice remaining coherent during intercalation. / by Dinesh Balachandran. / S.M.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Materials Science and Engineering, 2015. / 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 126-130). / Recycling provides a key strategy to move towards a more sustainable society by partially mitigating the impact of fast-growing material consumption. Recent advances in reprocessing technologies enable recyclers to incorporate low-quality secondary materials into higher quality finished products. Despite technological development, the use of these materials in the re-melting stage to produce final alloys is still limited. This thesis addresses this issue by raising the following question: given the complexity of the reprocessing operational environment, what is the most effective way to manage two-stage recycling operations to maximize the usage of low-quality secondary materials? This thesis answers this question for two systems: when outputs from the reprocessing stage can be delivered (1) as sows and (2) as liquid metals to the re-melting stage. In the first system, the main barrier to use of these materials is the highly variable quality of raw materials. This study suggests the use of data mining as a strategy to manage raw materials with uncertain quality using existing data from the recycling industry. A clustering analysis provides criteria for grouping raw materials by recognizing the pattern of varied compositions. This grouping (binning) strategy using the clustering analysis increases the homogeneity and distinctiveness of uncertain raw materials, allowing recyclers to increase their usage while maintaining minimum information about them. In the second system, significant energy cost can be saved by immediately incorporating reprocessed secondary raw materials as liquid metal into final alloy production. In this case, the coordination between the reprocessing stage and the re-melting stage is critical. This study suggests integrated production planning for two stages. The mathematical pooling problem is used to model two-stage recycling operations. Integrated planning across the two operations can adjust batch plans and design intermediate products by reflecting demand information of final products. This approach maximizes the use of intermediate products as liquid in the remelting stage and, therefore, lowers energy cost significantly. Both strategies are applied to industrial cases of aluminum recycling to explore the benefits and limitations. The results indicate the potential opportunity to significantly reduce material costs and to increase the use of undervalued secondary raw materials. / by Jiyoun Christina Chang. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2011. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 117-125). / Grain boundaries and their networks have a profound influence on the functional and structural properties of every class of polycrystalline materials and play a critical role in structural evolution and phase transformations. Recent experimental advances enable a full crystallographic characterization, including the boundary misorientation and inclination parameters, of grain boundaries. Despite these advances, a lack of appropriate analytical tools severely undermines our ability to analyze and exploit the full potential of the vast amounts of experimental data available to materials scientists. This is because the topology of the grain boundary space is unknown and even a well-studied part of the complete grain boundary space, the misorientation space, is relatively poorly understood. This thesis summarizes efforts to improve the representation of misorientation information and to understand the topology of the complete grain boundary space. First, the topology of the space of misorientations is discussed with a focus on the effect of symmetries on the minimum embedding dimensions in Euclidean space. This opens the door to a new method of representation of misorientation information in which grain boundaries can be uniquely colored by their misorientations. Second, conditions under which the topology of the grain boundary space has been resolved are presented. Resolving the topology of the complete grain boundary space not only facilitates statistical analysis of grain boundaries, but can also help describe the structure-property relationships of these interfaces. / by Srikanth Patala. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2006. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Includes bibliographical references (p. 183-194). / Solid oxide fuel cells are a potential electrical power source that is silent, efficient, modular, and capable of operating on a wide variety of fuels. Unfortunately, current technologies are severely limited in that they provide sufficient power output only at very high temperatures (>800°C). One reason for this is because the electrodes have very poor (and poorly understood) kinetics. The work described in this dissertation involves the microfabrication of model systems with triple phase boundary lengths that varied over an order of magnitude to systematically quantify and ultimately improve the kinetics of platinum electrodes on the surface of yttria stabilized zirconia electrolytes. Platinum electrodes with well controlled geometry were sputtered onto the surface of bulk YSZ and onto sputtered YSZ thin films. An unexpected result was found whereby YSZ films of composition Y0.09Zr0.91O2-x had an ionic conductivity remarkably enhanced by a factor of 20-30. This is attributed to the films exhibiting nanometric grain sizes and thereby stabilizing the cubic morphology at considerably lower yttrium levels than is normally needed. This metastable cubic phase is suspected of having reduced defect ordering. / (cont.) Grain boundary resistance, which in YSZ is normally due to impurities that segregate and block ionic transfer, was found to also be significantly reduced in YSZ films. The films had a specific grain boundary conductivity enhanced by a factor of 30-100 compared to the bulk polycrystalline sample. This was believed to be due to the very low impurity content of the film grain boundaries. Concerning the electrode polarization resistance, it was found that the electrodes placed on bulk standards and films deposited at high temperatures were on par with the best electrode conductance values from the literature. However, when the electrolyte surface was a film deposited at reduced temperature, the resistance decreased further by a factor of 300-500. The cause of this was revealed to be silicon contamination on the surfaces of the poorer-performing electrolytes. Triple phase boundary length-specific resistances as low as 3.7·104 O·cm at 378°C and 4.0·107 O·cm at 215°C were measured; these appear to be the lowest ever recorded. The measurements are possibly the first electrochemical characterization of nearly silicon-free YSZ surfaces. This study emphasizes the key role of chemical purity at the electrode-electrolyte interface. / (cont.) Photolithography alone is unlikely to give technologically useful triple phase boundary lengths. In an attempt to achieve the triple phase boundary lengths needed for a practical device, reactive co-sputtering was used to produce composite Pt-YSZ thin films with a bi-continuous network morphology and grain sizes on the order of 30 nm. Such intimate mixing of the electronic and ionic conducting phases created an effective mixed ionic-electronic conductor with the entire surface of the film electrochemically active to the electrode reaction. The best processing conditions resulted in electrodes with an area specific polarization resistance less than 500 O·cm2 at 400°C and, by extrapolation, 10 O·cm2 at 511°C and 1 O·cm2 at 608°C. These films may enable operation of a micro-solid oxide fuel cell at intermediate temperatures (400-500°C), and perhaps even lower temperatures with further microstructural optimization. / by Joshua L. Hertz. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2007. / Includes bibliographical references (p. 133-140). / Manipulation of the distribution of phonons inM a solid is important for both basic science and applications ranging from heat management to reduction of noise in electronic circuits and creating materials with superior acoustic and acousto-optical properties. This thesis explores hypersonic phononic crystals as means to achieve control over high frequency acoustic phonons. An integrated approach to fabrication, measurement and analysis of hypersonic phononic crystals with band gaps in the GHz range is presented. First, the phonon dispersion relation for one dimensional polymeric phononic crystals fabricated by coextrusion of a large number of poly(methylmethacrylate)/poly(carbonate) and poly(methylmethacrylate)/poly(ethylene terephthalate) bilayer pairs is investigated as a function of a lattice constant and composition using Brillouin light scattering and numerical simulations. This set of relatively simple multilayer structures represents an excellent platform to gain a basic understanding of phononic band gap phenomena. In addition, their in-plane phonon dispersion is used to extract information about the elastic constants and glass transition temperatures of individual nanolayers in a periodic multilayer arrangement. Next, two dimensional epoxy/air phononic crystals fabricated in a photoresist using interference lithography are studied. These structures are 2D single crystalline, enabling direction-resolved measurements of their phonon dispersion relation. As a result, the complete experimental phononic band diagram is obtained and correlated with numerical simulations. Finally, phononic properties of three dimensional elastomeric poly(dimethylsiloxane) crystals are investigated and the mechanical tunability of their dispersion relation is demonstrated. / (cont.) This set of structures forms the basis for understanding how to design and fabricate acoustic and acousto-optical devices with performance characteristics that can be adjusted dynamically during operation. The investigations described in this thesis demonstrate both theoretically and experimentally that 1D, 2D and 3D periodic submicron structures have complex phonon dispersion relations at GHz frequencies. As a result, these crystals can be used to manipulate the flow of random thermal phonons as well as externally generated acoustic waves resulting in novel acoustic and thermal properties. / by Taras Gorishnyy. / Ph.D.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2011. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 90-92). / In order to satisfy the growing demand for renewable resources as a supply of electricity, much effort is being placed toward the development of battery energy storage systems that can effectively interface these new sources with today's electric grid. To be competitive with the prices dictated by the sources currently in use, namely fossil fuels, these new systems must be able to deliver energy at a cost of about $100/kWh for the active materials. Several battery systems have been developed and that target has been slowly coming into focus. The liquid metal battery attempts to redefine the typical storage system by maintaining components in a molten state and seeks to do so using cheap, widely available materials. One of the biggest problems facing the future of the LMB is that of resource scarcity. There are several candidate materials that can satisfy the operating needs of an LMB, but not all are available in large enough quantities to meet the new LMB demand without significant impact to the supply/demand equilibrium and, ultimately, price. A detailed model was built to investigate the use of various metals in this new technology and measure the impact expected as a result of large-scale LMB adoption. The report explains why antimony is the most likely candidate for the positive electrode due to its large, mature market and the relatively high energy output compared to other candidates. Alkali and alkaline earth materials make excellent candidates for the negative electrode and the vast quantities available will be more than enough to support the LMB market many times over. The analysis has also revealed a substantial concern that is often overlooked in battery development: electrolyte costs. The current use of ultra-pure, anhydrous salts cannot be sustained if the LMB is to be profitable and competitive. Large savings can be anticipated from purchasing tonnage-quantity salts, but the only way to bring these costs down to reasonable levels is through the use of lower-purity materials. This study shows that a reasonable LMB can be built with an active materials cost of less than $62/kWh and total system cost of around $1,000/kWh for a 1 MW facility. In the most optimistic case, assuming electrolyte costs on par with those of current battery technologies, the total sytem cost can be reduced to under $300/kWh, much cheaper than many of the most recent technologies. / by Michael C. Parent. / S.M.