Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1987. / Bibliography: leaves 114-116. / by Yongun Lee. / M.S.

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 142-149). / III-V semiconducting nanostructures present a promising platform for the realization of advanced optoelectronic devices due to their superior intrinsic materials properties including direct band gap energies that span the visible light spectrum and high carrier mobilities. Additionally, the inherently high surface-to-volume ratio of nanostructures allows for the efficient relaxation of stress enabling the realization of defect free heterostructures between highly mismatched materials. As a result, nanostructures are being investigated as a route towards the direct integration of III-V materials on silicon substrates and as platforms for the fabrication of novel heterostructures not achievable in a thin film geometry. Due to their small size, however, many of the methods used to calculate stress and strain in 2D bulk systems are no longer valid as free surface effects allow for relaxation creating more complicated stress and strain fields. These inhomogeneous strain fields could have significant impacts on both device fabrication and operation. Therefore, it will be vital to develop techniques that can accurately predict and measure the stress and strain in individual nanostructures. In this thesis, we demonstrate how the combination of advanced transmission electron microscopy (TEM) and continuum modeling techniques can provide a quantitative understanding of the complex strain fields in nanostructures with high spatial resolutions. Using techniques such as convergent beam electron diffraction, nanobeam electron diffraction, and geometric phase analysis we quantify and map the strain fields in top-down fabricated InAlN/GaN high electron mobility transistor structures and GaAs/GaAsP core-shell nanowires grown by a particle-mediated vapor-liquid-solid mechanism. By comparing our experimental results to strain fields calculated by finite element analysis, we show that these techniques can provide quantitative strain information with spatial resolutions on the order of 1 nm. Our results highlight the importance of nanoscale characterization of strain in nanostructures and point to future opportunities for strain engineering to precisely tune the behavior and operation of these highly relevant structures. / by Eric James Jones. / Ph. D.

Carbon nanotubes (CNTs) are known to exhibit outstanding mechanical, electrical, thermal, and coupled electromechanical properties. CNTs can be employed towards the design of an innovative strain sensor with enhanced multifunctionality due to their load carrying capability, sensing properties, high thermal stability, and outstanding electrical conductivity. All these features indicate the prospect to use CNTs in a very wide range of applications, for instance, highly sensitive resistance-type strain/force sensors, wearable electronics, flexible microelectronic devices, robotic skins, and in-situ structural health monitoring. CNT-based strain sensors can be divided into two different types, the individual CNT- based strain sensors and the ensemble CNT-based strain sensors e.g. CNT/polymer nanocomposites and CNT thin films. In contrast, to individual CNT-based strain sensors with very high gauge factor (GF) e.g. ~3000, the ensemble CNT-based strain sensors exhibit very low GFs e.g. for a SWCNT thin film strain sensor, GF is ~1. This research discusses the mechanisms and the optimizing principles of a SWCNT thin film piezoresistive sensor, and provide an experimental validation of the numerical/analytical investigations. The dependence of the piezoresistivity on key parameters like alignment, network density, bundle diameter (effective tunneling area), and SWCNT length is studied. The tunneling effect is significant in SWCNT thin films showing higher degrees of alignment, due to greater inter-tube distances between the SWCNTs as compared to random oriented SWCNT thin films. It can be concluded that SWCNT thin films featuring higher alignment would have a higher GF. On the other hand, the use of sparse network density which comprises of aligned SWCNTs can as well intensify the tunneling effect which can result to a further increase in the GF. In addition, it is well-known that percolation is greatly influenced by the geometry of the nanotubes e.g. bundle diameter and length. A study on the influence of bundle diameter of SWCNTs on the piezoresistivity behavior of mechanically drawn SWCNT thin films showed the best performance with an improved GF of ~10 when compared to the randomly oriented SWCNT thin films with GF of ~1. The non-linear piezoresistivity of the mechanically drawn SWCNT thin films is considered to be the main mechanism behind the high strain sensitivity. Furthermore, information about the average length and length distribution is very essential when examining the influence of individual nanotube length on the strain sensitivity. With that in mind, we use our previously developed preparative ultracentrifuge method (PUM), and our newly developed gel electrophoresis and simultaneous Raman and photolumiscence spectroscopy (GEP-SRSPL) to characterize the average length and length distribution of SWCNTs respectively. / A Dissertation submitted to the Department of Industrial and Manufacturing Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2015. / September 28, 2015. / Alignment, Gauge Fcator, Network density, Piezoresistivity, Single Walled Carbon nanotubes, Strain sensor / Includes bibliographical references. / Tao Liu, Professor Directing Dissertation; Sachin Shanbhag, University Representative; Mei Zhang, Committee Member; Okenwa Okoli, Committee Member; William Oates, Committee Member.

In this dissertation, microwave chemistry is employed to synthesize a variety of different crystalline nanoparticles (NPs). This introduction will describe the structures, properties and applications of the NPs studied within the dissertation, with a main focus being on ligand sensitization for the goal of enhanced luminescence. The use of metal acetylacetonate complexes to make Europium (III) doped Ytrrium (Y₂O₃) NPs is explored, where the acetylacetonate acts both as a source of oxygen for the synthesis of Y₂O₃, as well as an organic chromophore acting as an "antenna" for the absorption of light and subsequent excitation transfer to the incorporated Europium (III) (Chapter 2). Other host materials are investigated by method of metal acetylacetonate decomposition to synthesize a variety of different nanospinels, having the general formula AB₂X₄, with sulfide variants made by decomposition of diethyldithiocarbamate, (Chapter 3). The antenna ligand thenoyltrifluoroacetone (tta), which is known to undergo a Dexter energy transfer (DET) mechanism to efficiently sensitize Europium (III) emission, is used to determine the distance of energy transfer in Europium (III) doped nanospinels by passivating the surface of the nanospinel with a tta (Chapter 4). A variety of ligands are explored in order to optimize the sensitization efficiency in relation to the difference in energy between the singlet and triplet levels of the ligands versus the ⁵D₀ and ⁵D₄ energy levels of Europium (III) and Terbium (III) respectively (Chapter 5). / A Dissertation submitted to the Department of Chemistry and Biochemistry in partial fulfillment of the Doctor of Philosophy. / Fall Semester 2015. / October 28, 2015. / Includes bibliographical references. / Geoffrey F. Strouse, Professor Directing Dissertation; William M. Landing, University Representative; Albert E. Stiegman, Committee Member; Gregory B. Dudley, Committee Member.

Colloidal semiconductor quantum dots (QDs) are inorganic nanocrystals that possess several unique photophysical properties, including tunable narrow emission and remarkable photo- and chemical stability. They have large surface areas, and thus can be decorated with large numbers and a variety of molecular vectors. These properties combined offer a potentially superior alternative to traditional organic fluorophore for advanced applications in bio-imaging and bio-sensing. Herein, our effort has centered on developing a series of metal coordinating ligands with controllable structures to modify the QD surfaces and construct biocompatible nanocrystals. The ligand architecture accounts for several factors: (i) variable coordination number, (ii) nature of the hydrophilic moiety, polyethylene glycol (PEG) or zwitterion, and (iii) versatility of end-reactive groups including amine, azide, carboxylic acid and aldehyde. The ligand design is combined with a newly developed photoligation strategy to promote the dispersion of luminescent QDs in buffer media. The dissertation is organized in six chapters: In chapter 1, we provide a brief introduction of the basic photophysical properties of QDs and the synthesis history for growing high quality semiconductor nanocrystals. We also present some of the most effective methods reported to date to prepare aqueous QD dispersions, discuss the effective chemical coupling strategies for conjugating biomolecules, and review the recent literatures that have used QD-bioconjugates for imaging and sensing purposes. In Chapter 2, we describe a novel photoligation strategy to promote the transfer of luminescent QDs from hydrophobic to hydrophilic media using lipic acid (LA)-based ligands. We also discusse the experimental conditions, mechanismfor in-situ ligand exchange and the generosity of the method towards the diverse functionality while maintaining the optical properties of the nanocrystals. In chapter 3, we present the design and synthesis of three sets of compact zwitterionic ligands comprising either one or two lipoic acid (LA) groups chemically linked to a zwitterion moiety. These ligands are then combined with the photoligation strategy to promote the phase transfer of QDs to buffer media. The high compactness and the stability of the nanocrystals over a broad range of conditions have been discussed.This chapter also highlights the conjugation of mCherry to the QD surface via metal-histidine coordination, as a proof-of-concept, to develop FRET-based sensors. In chapter 4, we detail a versatile strategy to prepare a series of poly (ethylene glycol) containing multicoordinating ligands optimized for the surface-functionalization of luminescent QDs and gold nanoparticles (AuNPs) alike. Our chemical design relies on the modification of chiral L-aspartic acid precursor, and the advantages of using aminoacid combined with lipoic acid and reactive PEG moieties have been discussed. Nonetheless, the two sets of ligands: bis(LA)-PEG-FN and LA-(PEG-FN)₂ described here are compatible with photoligation strategy to yield hydrophilic, colloidally stable and reactive nanoparticles (QDs and AuNPs). In chapter 5, we discuss the preparation of hydrophilic QDs with intact azide (-N3) and aldehyde (-CHO) bio-orthorgonal functionalities on their surfaces. Strain promoted click chemistry and hydrazine ligation will be discussed to illustrate the orthogonality of two reactive groups, azide and aldehyde. Additionally, we demonstrate an optical method to extract the number of reactive -CHO groups per QD and finally estimate the total number of ligands bound to each QD for a few distinct size nanocrystals. / A Dissertation submitted to the Department of Chemistry and Biochemistry in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2015. / November 5, 2015. / Biocompatible, Bio-orthogonal, Ligand design, Photoligation, Quantum dots, Surface modification / Includes bibliographical references. / Hedi Mattoussi, Professor Directing Dissertation; Hengli Tang, University Representative; Gregory B. Dudley, Committee Member; Joseph B. Schlenoff, Committee Member.

In many composite materials, rigid fibers are distributed throughout the material to tune the mechanical, thermal, and electric properties of the composite. The orientation and distribution of the fibers play a critical role in the properties of the composite. Many composites are processed as a liquid molten suspension of fibers and then solidified, holding the fibers in place. Once the fiber orientations are known, theoretical models exist that can predict properties of the composite.Modeling the suspended fibers in the liquid state is important because their ultimate configuration depends strongly on the flow history during the molten processing. Continuum models, such as the Folgar-Tucker model, predict the evolution of the fibers’ orientation in a fluid. These models are limited in several ways. First, they require empirical constants and closure relations that must be determined a priori, either by experiments or detailed computer simulations. Second, they assume that all the fibers are slender bodies of uniform length. Lastly, these methods break down for concentrated suspensions. For these reasons, it is desirable in certain situations to model the movement of individual fibers explicitly. This dissertation builds upon recent advances in boundary integral equations to develop a robust, accurate, and stable method that simulates fibers of arbitrary shape in a planar flow. In any method that explicitly models the individual fiber motion, care must be taken to ensure numerical errors do not cause the fibers to overlap. To maintain fiber separation, a repulsion force and torque are added when required. This repulsion force is free of tuning parameters and is determined by solving a sequence of linear complementarity problems to ensure that the configuration does not have any overlap between fibers. Numerical experiments demonstrate the stability of the method for concentrated suspensions. / A Dissertation submitted to the Department of Scientific Computing in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester 2018. / July 16, 2018. / Boundary Integral Equations, Complementarity Problems, Rigid Particle Suspensions / Includes bibliographical references. / Bryan Quaife, Professor Co-Directing Dissertation; Sachin Shanbhag, Professor Co-Directing Dissertation; Nick Cogan, University Representative; Chen Huang, Committee Member; Nick Moore, Committee Member.

Vast information on existing crystal structures, which is available through the large open-access and commercialized repositories of crystallographic data, provides an excellent starting point for the implementation of deep learning techniques to the discovery of hidden relationships that might be contained in such large datasets. The machine learning algorithm can be thought of as a fitting procedure for a complicated heuristic model using a large amount of data.1-2 This model is later tested to estimate its ability to generalize to unknown crystal structures in a holdout set, i.e. its predictive ability. Herein, we describe a neural network model trained to predict the likelihood of chemical elements adopting different topologies of atomic sites in known crystal structures. The neural network is shown examples of topologies from known crystal structures and trained to predict the element that adopted that topology. We apply the trained model to predict possible compositions of unknown compounds that might be pursued by a synthetic chemist. We demonstrate that the deep neural network is capable of automatically “discovering” relevant descriptors from high-dimensional “raw representations” of the crystallographic data. Since the input data contain purely geometrical and topological information, any chemical knowledge residing within the neural network output must have been learned during training, and thus was “discovered”. The neural network’s learned representation of local topology shows evidence of known geometric and chemical trends not explicitly provided to the network during training. / A Dissertation submitted to the Department of Chemistry and Biochemistry in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester 2018. / July 10, 2018. / Crystallography, Deep Learning / Includes bibliographical references. / Mykhailo Shatruk, Professor Directing Dissertation; Adrian Gheorghe Barbu, University Representative; Albert E. De Prince, Committee Member; Susan E. Latturner, Committee Member.

Cells sense and respond to topographical cues in their microenvironment that influence growth, development, and migration. Cell migration and outgrowth assays have been used to study cellular movement or changes in cellular morphology and topography. Such assays are promising tools in drug discovery, especially when implemented with high-throughput and high-content imaging systems. These techniques have also been useful for screening and analyzing the effect of different compounds on neurite outgrowth and topography which in turn may lead to the discovery of beneficial targets for regeneration of nervous tissue. Typically, high-throughput screening of large chemical libraries is employed during the early stages of discovering new drug entities. However, these screening assays do not utilize different topographical surfaces. Many common techniques such as the scratch wound assay are limited in their compatibility with patterned surfaces. Therefore, there is a need to develop novel technologies capable of identifying potentially therapeutic compounds in early stage of drug discovery processes that can regulate cell behaviors and are not limited in their throughput and compatibility with patterned surfaces. A potentially scalable approach is the “fence” assay in which cells are cultured on topographical surfaces which are partially covered by a removable barrier. Upon removal of the barrier, cells are free to spread and migrate on the freshly uncovered topographies. In this thesis, a novel technique called evaporative edge lithography (EEL) is demonstrated as an approach to miniaturize the fence assay and can be used for high-throughput screening (HTS) in early stages of drug discovery. Furthermore, EEL is a new method to fabricate lipid-based drug delivery microarrays. Lipid multilayer micro-patterns offer a promising approach to applications such as drug screening and biosensing that require well defined patterns and fluidity. It is shown in this thesis that the factors that govern stability and instability of lipid multilayer nanostructures upon immersion using fluorescence microscopy and observed the following four mechanisms of lipid multilayer instability and strategies are derived to control immersion stability based on these findings: (1) Dissolution by the air/water interface; (2) Disruption by shearing from flowing solution; (3) Spreading at the solid-liquid interface; (4) Diffusion into solution. Based on these studies, a lipid multilayer microarray was developed that is suitable for cell-based assays without detectible cross-contamination by culturing cells on lipid patterns. It is shown in this thesis that this assay is compatible with poorly soluble lipophilic drug compounds that pose a challenge for HTS microarray assays. EEL was demonstrated for topographically patterned surfaces for screening compounds on adherent cells. Lipophilic compounds including docetaxel and BFA were screened using EEL with cultured HeLa cells to test if migration is affected and can be quantified with this approach. These results indicate that docetaxel and BFA were delivered locally into cells locally from surface supported lipid films and significantly inhibited cellular migration. Subsequently, EEL was used to screen docetaxel on cultured primary olfactory bulb neuronal cells to test the effect on neurite outgrowth. EEL is a novel approach that allows delivery and subsequent study of the effects of poorly water-soluble drugs on cell migration as well as in vitro screening of different drugs for their effects on cell structures and functions. In addition, this migration assay is a scalable and promising approach for high throughput drug screening microarrays since multiple drug compounds at different dosages can be screened simultaneously on the same surface. This work will advance future studies in developing a portable assay capable of screening lipophilic cancer and neurotropic compounds for topographically-driven cell outgrowth and migration. / A Dissertation submitted to the Department of Biological Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester 2018. / April 5, 2018. / cell migration, drug screening, evaporative edge lithography, lipid multilayers, lipid patterning, neurite outgrowth / Includes bibliographical references. / Steven Lenhert, Professor Directing Dissertation; Jingjiao Guan, University Representative; Kathryn Jones, Committee Member; Thomas Keller, Committee Member; Jonathan Dennis, Committee Member.

For 10-15 years, the Powder-In-Tube (PIT) process has been one of the leading manufacturing methods for producing the highest critical current density (Jc) Nb₃Sn wires for at small effective filament diameter (deff), both required for future applications in high energy physics. Since Nb₃Sn first became commercially available in the 1960's, non-Cu Jc values have steadily climbed until a plateau was reached at about 3,000~A/mm (12~T, 4.2~K) in the year 2000. Comprehensive analysis of recent wires suggests that both PIT and the other high Jc wire design, Rod Restack Process (RRP), have yet to achieve their maximum potential as high Jc conductors. Currently, PIT wires obtain a maximum Jc(12~T, 4.2~K) of about 2700~A/mm2 and do so by converting up to 60% of the non-Cu cross section into superconducting Nb₃Sn. However, about a quarter of this volume fraction is made of large grains of Nb₃Sn which are too large or otherwise disconnected to carry current in transport, wasting both real estate, as well as Sn and Nb that would be better used to make the desired small grain A15. The most recent RRP wires typically achieve Jc values of around 3000~A/mm2 by also converting about 60% of the non-Cu cross section into A15, however nearly all of that has the desired small grain morphology with high vortex pinning, ideal for current transport. Studies at the Applied Superconductivity Center have shown that one route to improvement for both wires may be in controlling the formation of intermediate phases which form before Nb₃Sn . An intermetallic Nb-Cu-Sn, commonly referred to as Nausite, is considered responsible for the formation of the undesirable large grains. We studied the phase evolution in PIT Nb₃Sn from the starting powder mixture at room temperature up to 690°C to better understand what role Nausite ((Nb0.75Cu0.25)Sn₂) actually plays in forming large grain A15 with the goal of preventing its formation and making better use of the Sn available to form the desired small grain A15 morphology. After heat treatment, all wires were imaged in an SEM and then processed through digital image analysis software, extracting area fractions of each phase and their morphology. For heat treatments which showed interesting metallographic results, additional measurements were made by transport, resistivity, magnetization, and/or heat capacity to develop a complete picture of how the microstructure affects critical wire properties. Based on these results, novel heat treatments were developed and demonstrated our ability to reduce the undesired large grain A15 while simultaneously producing more current-carrying small grain A15, increasing the ratio of small:large grain A15 from 3.0 to 3.8. Another possible path to improvement is to reduce the non-uniform deformation incurred during wire fabrication. A PIT Nb₃Sn wire begins as a mono-filament consisting of a thick Nb7.5wt%Ta tube clad in high-purity Cu, inscribed with a Cu sleeve, and filled with a Sn-rich NbSn₂ powder. The external Cu cladding will later provide a low resistance normal conducting path around superconducting filaments, a necessity for magnet stability. The final wire diameter is between 0.7-1.25~mm with 156 or 192 filaments, whose diameters are 33-50~μm, organized into 6-7 concentric rings. Through advanced digital image analysis software, we can extract geometric information which describes how the wire and filaments deform from their nominally circular shape, becoming elliptical or otherwise having non-uniform deformation which can be detrimental to wire properties. We found that the non-uniform deformation incurred during wire fabrication can degrade the wire performance. The most severe effect is caused by the different deformation rates of the Nb-Ta tube compared to its powder core, which leads to the Sn-rich core drifting from the center of the Nb-Ta tube, leading to an uneven A15 reaction front. This is referred to here as 'centroid drift'. In PIT wires, the diffusion barrier must be consumed to form A15 while still leaving a thin, protective annulus behind to protect the Cu. Centroid drift then causes a large inefficiency as it creates a thick and thin side of diffusion barrier, the thin side limiting the reaction if Sn leaks are to be prevented, while the thick side becomes wasted Nb-Ta. Up to 30% of the final non-Cu cross-section remains as unused diffusion barrier. When Sn leaks out of the filaments it increases the resistivity of the Cu stabilizer, lowering the Residual Resistance Ratio (RRR). A high RRR is required for magnet stability, and such Sn leaks can be detrimental to magnet performance. In addition, we found that filaments farther from the center of the wire tend to be those with highest centroid drift, and they are also the most susceptible to leaking Sn. Moreover, we observed that in leaks severe enough to produce a Kirkendall void, the A15 volume is also reduced. By comparison, RRP type wires manage a similar reaction with less than 10% residual barrier in the non-Cu cross section. Recently, Bruker EAS, the manufacturer of PIT Nb₃Sn wires, developed a new wire design which added a bundle barrier around the filament pack to contain Sn leaks and maintain a high RRR, as well as increasing the Sn content in the powder core to produce more A15. We believe that by improving deformation properties and optimizing new heat treatments to account for the higher Sn content, Jc can be substantially enhanced while maintaining RRR at small filament diameters. / A Dissertation submitted to the Department of Mechanical Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester 2018. / April 16, 2018. / CERN, magnets, nb3sn, powder in tube, superconductivity, wires / Includes bibliographical references. / David C. Larbalestier, Professor Directing Thesis; Jennifer Proffitt, University Representative; Fumitake Kametani, Committee Member; Emmanuel Collins, Committee Member; Chiara Tarantini, Committee Member; Peter J. Lee, Committee Member.

Halide perovskite materials have emerged in the past few years as promising materials in the absorption layer of photovoltaic cells and new emissive materials for use in light emitting diodes (LEDs). This is due to their rapidly increasing efficiencies and brightness. In photovoltaic applications they show promise to lower cost and improve efficiency of photovoltaic cells. Their low temperature processability also may lead to interesting new applications in existing solar cell technologies. In LED applications, they exhibit other desirable properties such as color tunability, simple device structures, and facile processability. However, a common problem that is observed in perovskite thin films is a hysteresis in their I-V characteristics, and short device lifetimes. It is hypothesized this is due to ion migration within the crystal and along the grain boundaries between crystals. This thesis addresses this issue by exploring methods to restrict ionic motion. One highly promising method was controlling nucleation to reduce the grain boundary density in the perovskite thin films. A deterministic nucleation process was developed using standard lithography techniques to prepattern a substrate followed by solution processing of the halide perovskite layer. It was found the grain size, grain boundary density, and final crystal shape could be well controlled using this process. In addition, it was found the hysteresis behavior was well controlled, and the stability of the final film was increased due to lower grain boundary density. In addition, further methods to restrict ionic motion were explored using Ruddlesden-Popper perovskites that form a quasi 2D structure. These perovskites were examined and characterized due to their ability to restrict ionic motion within the perovskite crystal. These perovskites also allowed for further flexibility in tuning device electrical and optical properties and offered greater stability compared to their 3D counterparts. / A Dissertation submitted to the Program in Materials Science and Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2018. / November 5, 2018. / 2D Materials, Grain Boundaries, Halide Perovskites, Optoelectronics / Includes bibliographical references. / Zhibin Yu, Professor Directing Dissertation; John Telotte, University Representative; Kenneth Hanson, Committee Member; Zhiyong Richard Liang, Committee Member; Yan-Yan Hu, Committee Member.