Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2003. / Includes bibliographical references (leaves 107-118). / Angiogenesis plays a key role in tumor growth. The dependency of tumors on angiogenesis has rendered it a promising therapeutic target. However, to date, this promise has gone unfulfilled in the clinic, suggesting that the current understanding of angiogenesis is insufficient. The objective of this dissertation is to quantitatively analyze the effects of systemic biochemical and cellular contributions as well as local mechanical influences on angiogenesis using a combination of theoretical and experimental approaches. A model of the balance between angiogenic stimulators and inhibitors is developed to assess the effects of biochemical factors produced by the primary tumor on angiogenesis both locally and remotely. The model quantitatively describes how primary tumors can suppress metastases and provides a framework for assessing the conditions under which this may occur. The model also predicts a disruption of the balance between angiogenic and anti-angiogenic factors within the primary tumor that may result in distinct regions of angiogenesis stimulation and suppression, offering a new hypothesis for the experimentally observed formation of central necrosis. Based on these predictions of angiogenic activity within the primary tumor, a model of the contribution of endothelial progenitor cells to tumor angiogenesis is advanced. The model accurately captures the salient features of tumor growth and angiogenesis and predicts that endothelial progenitor cells make a significant contribution to tumor growth and angiogenesis. Model simulations of several anti-angiogenic therapeutic strategies indicate that effectively targeting pathways affecting both vessel wall-associated endothelial cells and circulating endothelial progenitor cells leads to improved outcome as compared to targeting either pathway alone. / (cont.) The hypothesis that unusual tumor blood vessel morphology is caused by solid mechanical interactions is investigated experimentally. Blood vessels in tumors receiving treatment cytotoxic to the tumor cells are significantly more open and circular than those in tumors receiving no treatment. The interaction and coupling between these biochemical, cellular, and mechanical influences on tumor angiogenesis and their potential biological and clinical implications are discussed. / Brian R. Stoll. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2005. / Includes bibliographical references. / Segmented polyurethane elastomers containing additional ordered structures within the hard or soft domains were developed to mimic the hierarchical structure and superior properties observed in spider silk fibers. The silk's toughness is related to a fiber morphology that includes P-pleated crystalline sheets within an amorphous matrix, as well as an additional interphase with an orientation and mobility between that of the two microphases. In the polyurethane mimics, bulky aromatic diisocyanates were incorporated between aliphatic hexamethylene diisocyanate (HDI) hard segments and poly(tetramethylene oxide) (PTMO) soft segments, to enhance the size and orientation of the interphase. The mixture of diisocyanates reduces the crystallinity of the HDI hard segments, allowing the polyurethane to form more well-organized domains observed by AFM imaging. The more interconnected hard domains allow the elastomers to deform to higher elongations and absorb more energy without a decrease of initial modulus. Shearing of the hydrogen-bonded hard domains orients the hard blocks at a preferred tilt angle of ±20⁰ from the strain direction during tensile deformation. / (cont.) While the average spacing of hard domains increases during deformation, the spacing of hard domains aligned with the strain decreases, and the spacing of hard domains at the preferred tilt angle remains constant. Strain-induced crystallization of the PTMO soft segments was observed in all samples; however, hard segments with mixed diisocyanates exhibited non-crystalline alignment of the hard domains. Several polyurethane nanocomposite structures were also created using particles that preferentially associate with hard or soft segments. HDI-PTMO polyurethane/Laponite nanocomposites provided modest mechanical property improvements (80% increase in modulus and 15% increase in toughness) without any loss of extensibility. The Laponite discs exhibited an exfoliated structure, associating with and reinforcing the hydrophilic polyurethane hard segments. HDI-PTMO polyurethane/MQ siloxane resin nanocomposites also exhibited particle association with the hard segments, providing a 60% increase in modulus with a small loss of toughness. / (cont.) However, composites of isobutyl-POSS dispersed in polyurethanes with mixed hard segments exhibited formation of POSS crystals associated with the soft segments at all loadings, resulting in tensile failure at strains 80-100% lower than the pure polyurethane. / by Gregory Stewart Pollock. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical 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. / by Vasileios Theodoros Dendroulakis. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1996. / Includes bibliographical references (leaves 136-142). / by Vivek Mohindra. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Chemical Engineering, 2016. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 175-182). / Recent advances in theory and practice, have introduced a wide variety of tools from machine learning that can be applied to data intensive chemical engineering problems. This thesis covers applications of statistical learning spanning a range of relative importance of data versus existing detailed theory. In each application, the quantity and quality of data available from experimental systems are used in conjunction with an understanding of the theoretical physical laws governing system behavior to the extent they are available. A detailed generative parametric model for optical spectra of multicomponent mixtures is introduced. The application of interest is the quantification of uncertainty associated with estimating the relative abundance of mixtures of carbon nanotubes in solution. This work describes a detailed analysis of sources of uncertainty in estimation of relative abundance of chemical species in solution from optical spectroscopy. In particular, the quantification of uncertainty in mixtures with parametric uncertainty in pure component spectra is addressed. Markov Chain Monte Carlo methods are utilized to quantify uncertainty in these situations and the inaccuracy and potential for error in simpler methods is demonstrated. Strategies to improve estimation accuracy and reduce uncertainty in practical experimental situations are developed including when multiple measurements are available and with sequential data. The utilization of computational Bayesian inference in chemometric problems shows great promise in a wide variety of practical experimental applications. A related deconvolution problem is addressed in which a detailed physical model is not available, but the objective of analysis is to map from a measured vector valued signal to a sum of an unknown number of discrete contributions. The data analyzed in this application is electrical signals generated from a free surface electro-spinning apparatus. In this information poor system, MAP estimation is used to reduce the variance in estimates of the physical parameters of interest. The formulation of the estimation problem in a probabilistic context allows for the introduction of prior knowledge to compensate for a high dimensional ill-conditioned inverse problem. The estimates from this work are used to develop a productivity model expanding on previous work and showing how the uncertainty from estimation impacts system understanding. A new machine learning based method for monitoring for anomalous behavior in production oil wells is reported. The method entails a transformation of the available time series of measurements into a high-dimensional feature space representation. This transformation yields results which can be treated as static independent measurements. A new method for feature selection in one-class classification problems is developed based on approximate knowledge of the state of the system. An extension of features space transformation methods on time series data is introduced to handle multivariate data in large computationally burdensome domains by using sparse feature extraction methods. As a whole these projects demonstrate the application of modern statistical modeling methods, to achieve superior results in data driven chemical engineering challenges. / by Mark Christopher Molaro. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1994. / Includes bibliographical references (leaves 205-210). / by Juan Alfredo Gomez Giammattei. / Ph.D.

Thesis (M.S.) Massachusetts Institute of Technology. Dept. of Chemical Engineering, 1951. / Bibliography: leaves 142-144. / by Edward Rolfe. / M.S.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, September 2004. / Includes bibliographical references. / (cont.) The DEM simulation was used to simulate the relative importance of cohesion and friction. For angle of repose simulations, increasing the cohesion increased the final angle in a consistent, linear fashion. Increasing the interparticle friction coefficient increased the final angle up to a critical friction coefficient. For the range of two dimensional simulations with the particles shaped as discs, this critical friction coefficient was about 0.30. Above this threshold, increasing the interparticle friction had no impact on the angle of repose. This suggests that for most pharmaceutical powders, the cohesion and shape are the most important particle properties. Case studies relating interparticle adhesion to labscale powder performance (flow and blends) were completed with two active pharmaceutical ingredients and a number of excipients. In all of the flow cases, the rank order of interparticle or intermaterial adhesion forces measured with AFM was exactly predictive of the rank order of ease of flow. Similarly, in all of the blending case studies, the rank order of adhesion force between the active pharmaceutical ingredient and the excipient was exactly predictive of the rank order stability of blends of the materials. The blend stability was quantified using an on-line, nondestructive, noninvasive light induced fluorescence (LIF) instrument. Separately, the LIF instrument was used to estimate the content of fluorescent drugs (caffeine and triamterene) in tablets by measuring the surface fluorescence. A theoretical description of the accuracy of the surface measurement to correctly estimate the total content in a tablet ... / The flow of fine powder has not been fully understood nor has it been predictable by current simulation techniques. With the use of atomic force microscopy (AFM), interparticle forces of attraction and interparticle friction forces between fine particles have been measured with unprecedented success. In a novel coupling of technologies, the microscopic interparticle force data was used in a discrete element model (DEM) simulation to predict the bulk powder flow in multiple geometries. Excellent agreement between the simulated angles of repose for both glass beads and a microcrystalline cellulose pharmaceutical excipient was obtained. Qualitative agreement between simulation and experiment of flow in a rotating box was also obtained for these materials. The ability to measure interparticle friction forces for non-spherical particles was a novel development. The presence of surface asperities was shown to be directly responsible for changes in friction force. Interparticle friction coefficients were established by varying the applied load between particles. The average interparticle friction between the cellulose particles was 0.44, double the value for glass beads. For both particle types, the population of measured friction coefficients was well represented by a normal distribution. Interparticle friction was also measured between lactose particles. However, the surfaces of the lactose particles smoothed under high applied loads and the measured friction coefficients were lower (an average value of 0.26) than the cellulose particles. The similar friction coefficients for glass beads and smooth lactose particles further suggest that the interparticle friction was strongly dependent upon the surface asperities. / by Reuben Dumont Domike. / Ph.D.

Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1970. / Vita. / Bibliography: leaves 230-233. / by Edward S. Matulevicius. / Sc.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2009. / Cataloged from PDF version of thesis. / Includes bibliographical references. / The process of charging a polymer solution to draw a filament is known as electrospinning. Electrospinning is capable of producing a continuously depositing jet of controllable micron and sub-micron diameters. As fiber deposits, a nonwoven mat of randomly oriented fibers in two dimensions is generated. The mat is mechanically robust and suitable for a wide variety of applications due to its high surface area to mass ratio, controllable size scale and surface chemistry, and large void fraction. The number of publications on the topic of electrospinning continues to grow exponentially, as the experimental apparatus is relatively inexpensive to assemble and 1 mm thick fiber mats can be generated in as little as 2 hours. Many publications have focused on potential applications or the processing of specific materials. Some publications have reported on the hydrodynamics and physics of the electrospinning process, leading to an increased control of fiber diameter and morphology. One area that remains relatively unexplored is pore diameter and porosity within the fiber mat. The present work explores characterizing and controlling void space in electrospun materials and the use of these materials in the field of tissue engineering. Characterization and prediction of overall void fraction and individual pore diameter is first addressed. Mercury porosimetry was used to establish two physical parameters useful in electrospinning applications: average pore diameter and peak pore diameter. Average pore diameter refers to the volume-weighted average determined by the volumetric profile. Peak pore diameter is the pore diameter at which the largest amount of void volume becomes accessible. / (cont.) The accuracy of mercury porosimetry was also addressed, leading to a method of data correction for buckling of pores under the significant pressure generated by mercury porosimetry. Having characterized and predicted the void statistics for as-spun materials, the second portion of this work sought to use post-processing techniques to alter the effective pore diameter. Two components - poly(E-caprolactone) and poly(ethylene oxide) - were electrospun together, either from a common polymer solution or adjacent fluid jets on to a common target. Water was used to selectively remove the poly(ethylene oxide) component in both systems, with vastly different results. Mats electrospun from a common solution saw an increasing reduction in the void diameter with increasing poly(ethylene oxide) removal due to poly(E-caprolactone) chain rearrangement and contraction of the polymer fibers, up to a pore diameter reduction of 80%. Mats produced by the dual jet method saw both an increase and decrease of the effective pore diameter depending on processing conditions. These experiments represent the first steps by researchers to specifically tailor pore diameter independent of porosity or fiber diameter. The final portion of this thesis deals with the use of electrospun materials as 3-Dimensional tissue engineering scaffolds. An effective perfusion technique was developed for the seeding and infiltration of cells into multiple electrospun mats simultaneously, with 100% efficiency. This represents an enormous advantage over conventional seeding methods. / (cont.) Human Dermal Fibroblasts were seeded into scaffolds of drastically varied fiber diameter (300nm to 8 [mu]m) and morphology (beaded vs. uniform diameter). Despite cells spreading along large fibers instead of developing multiple attachment points between fibers, cell proliferation was greatest in scaffolds with pore diameters greater than 6 [mu]m. At the same time, mats with pore diameters less than 12 pm observed the greatest extracellular matrix growth. Additional investigation would be well-served to determine optimal parameters for cell dispersion and reproduction throughout the electrospun template across multiple cell phenotypes. / by Joseph Lenning Lowery. / Ph.D.