Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2008. / Includes bibliographical references (p. 191-200). / Dissipative particle dynamics (DPD) is a mesoscale simulation technique which uses soft potentials between large particles to reproduce liquid behavior. In form, DPD is similar to molecular dynamics, as all matter is represented by point particles which interact with each other via, pairwise forces. The method was first introduced in the early 1990's, and has since undergone a number of refinements which have put it on a firm thermodynamic footing. DPD is notable for the flexibility it presents the modeler for building complex fluid systems. DPD has been used to study simple molecular liquids, polymer and colloid solutions, and phase behavior of block copolymer melts. Recently, a number of workers have used DPD to study the flow of polymer solutions in various geometries such as microchannels, pores, and sudden contractions. While these types of flows are well-suited to DPD's relative strengths, an important step has been skipped. Before the results of these complex flows can be accepted, it is necessary to demonstrate that the rheological predictions made by DPD are generally reliable. The principle aim of this thesis is to demonstrate that the rheology of polymer solutions can be simulated successfully with DPD. The rheology of a solution of DPD dumbbells using a FENE spring force law is studied in the first part of this thesis via simulation of steady shear flow and steady planar elongational flow. The rheological results are compared to dilute Brownian dynamics simulations of the same FENE dumbbell model. The level of coarse-graining of the DPD fluid is varied by changing the length of the DPD dumbbell relative to the particle size, while maintaining a constant extensibility parameter. Broadly speaking, the viscosity, first normal stress coefficient, and dumbbell extension in shear flow calculated with DPD are in agreement with the BD results. The two methods are not perfectly alike however, and two systematic differences between the DPD and BD results are observed. An excluded volume effect which occurs naturally in DPD and is not present in the BD simulations results in elevated viscosity and dumbbell extension in the zero-shear-rate regime. / (cont.) The effect is more powerful in DPD dumbbells which are more coarse-grained. At high shear rates in the power-law regime, DPD systematically overpredicts the rate of shear-thinning, with the greatest deviation occurring in the most coarse-grained dumbbells. This is hypothesized to be a result of hydrodynamic interaction which comes naturally out of DPD's explicit treatment of the solvent. The HI effect is analyzed using the Giesekus anisotropic drag tensor. Shortly after its introduction, the complaint was made that DPD's dynamic results are suspect because it has a very low, gas-like Schmidt number, meaning that momentum and mass are transported through the DPD medium at similar rates. This is in contrast with physical liquids, which have large Schmidt numbers. The use of the Lowe-Anderson formulation of DPD allows the Schmidt number of a solution to be varied for the same polymer model. Shear flow simulations of identical dumbbells under different Schmidt number conditions give results in excellent agreement with each other, indicating that the Schmidt number is not an important factor in determining polymer rheology with DPD. Steady planar elongational flow is simulated for the first time in DPD using the Kraynik and Reinelt boundary conditions, which are periodic in both space and time, allowing for simulations of planar elongational flow for an unlimited period of time. The planar elongational flow results of FENE dumbbells are also in agreement with BD, butshow the same systematic deviations observed in shear flow. The second portion of this thesis examines a more complex polymer solution using DPD, with simulations of semidilute solutions of longer N = 20 bead-spring chain polymers undergoing shear and planar elongational flow. In addition to concentration effects, the importance of the solvent quality is also examined with simulations of polymer solutions in both good and theta solvents. In order to capture concentration dependency, a spring -spring repulsion force is added to the DPD model to prevent polymer springs from passing though each other. A strong concentration dependence on the longest relaxation time is observed. / (cont.) In planar elongational flow, each solution goes through a coil-stretch transition at the theoretically predicted strain rate De = 0.5. In shear flow, the rheological results are in qualitative agreement with theory, showing a plateau at low De, and a transition into a shear-thinning regime beginning at De = 1. While the planar elongational flow results show clear dependence on the solution relaxation time, the shear results show a mixed dependence on the overall solution relaxation time, which reflects the concentration dependence, and the relaxation rate of an isolated chain, suggesting that only some aspects of the shear rheology are affected by the concentration. The conclusion of this thesis is that DPD is able to faithfully reproduce reliable rheological behavior with bead-spring polymer models. We find however, that the computational costs associated with the explicit simulation of the solvent put DPD at a disadvantage for systematic rheology studies when compared to Brownian dynamics. The high costs of the spring-spring repulsion force implementation are particularly limiting. In complex systems where DPD's natural flexibility in molecular architecture and chemistry make it the best choice, rheological results can now be accepted with more confidence. / by Theis Forman Clarke. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2012. / Cataloged from PDF version of thesis. / Includes bibliographical references. / The past few years have introduced a flurry of interest over renewable energy sources. Biofuels have gained attention as renewable alternatives to liquid transportation fuels. Microbial platforms for biofuel production have become an attractive option for this purpose, mitigating numerous challenges found in crop-based production. Towards this end, metabolic engineering has established itself as an enabling technology for biofuels development. In this work we investigate the strategies of metabolic engineering for developing a biodiesel production platform, utilizing the oleaginous yeast Yarrowia lipolytica as the host organism. We establish new genetic tools for engineering Y. lipolytica beginning with an expression vector utilizing the genetic features from translation elongation factor 1-a (TEF). Additionally, a complementary plasmid was developed allowing for multiple plasmid integration. Bioinformatics analysis of intronic genes in hemiascomycetous yeast also identified relationships between functional pathways and intron enrichment, chronicling the evolutionary journey of yeast species. Next gene targets were examined within the lipid synthesis pathway: acetyl-coA carboxylase (ACC), delta9-desaturase (D9), ATP citrate lyase (ACL), and diacylglycerol acyltransferase (DGA). A combinatorial investigation revealed the order of contribution to lipid overproduction (from strongest to weakest): DGA, ACC, D9, ACL. Scale-up batch fermentation of selected strains revealed exceptionally high lipid accumulation and yield. These results demonstrate the balance between cellular growth and lipid production which is being modified through these genetic manipulations. We next explored utilization of alternative substrates to expand the capabilities and utility of Y. lipolytica. For xylose, a prevalent substrate in cellulosic feedstocks, expression of the redox pathway from Scheffersomyces stipitis and adaptation led to successful substrate utilization. Through the use of cofermentation, growth and productivity on xylose was improved dramatically with xylose-to-lipids conversion successfully demonstrated. For acetate, a potentially useful substrate for electrofuel production, lipid production using our strongest performing strain resulted in high lipid accumulation and yield. From this study, metabolic engineering of Y. lipolytica was successfully used to achieve exceptional lipid overproduction from a variety of substrates. Our genetic tools and recombinant strains establish a strong platform for the study and development of microbial processes for the production of biofuels. / by Mitchell Tai. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, February 2010. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 178-210). / In this thesis, metabolic routes were developed for the production of hydroxyacids and their lactones in multiple microbial systems. These compounds see widespread use in the production of pharmaceuticals, polymers, and fine chiral intermediates. First in this thesis, strategies and tools for metabolic pathway design are discussed. This is followed by the descriptions of and data for each microbial production system. The compounds produced in this thesis and their highest titers obtained are shown below: 3-Hydroxybutyrate (3HB) 2.9 g/L 3-Hydroxyvalerate (3HV) 5.3 g/L 4-Hydroxyvalerate (4HV) 27.1 g/L 4-Valerolactone (4VL) 8.2 g/L 3,4-Dihydroxybutyrate (DHBA) 3.2 g/L 3-Hydroxybutyrolactone (3-HBL) 2.2 g/L The production of the two hydroxyvalerates was accomplished through the reduction of the renewable substrate levulinate in Pseudomonas putida by endogenous host enzymes followed by the liberation of the hydroxyvalerate product through the recombinant expression of thioesterase B (tesB). The production of 4VL was accomplished from levulinate by adding the lactonase paraoxonase I (PON1) to the P. putida hydroxyvalerate production system. Because 4VL was found to exist in a pH-dependent equilibrium with 4HV, the lactonase was expressed extracytosolically in acidic media to achieve significant titers of 4VL. The addition of a second resin phase to 4VL-producing cultures with a high affinity for 4VL substantially enhanced lactone production. 3HB, 3HV, DHBA, and 3-HBL were all produced in Escherichia coli through the expression of an acetoacetyl-CoA thiolase (thil, bktB, or phaA), a 3-hydroxybutyryl-CoA reductase (phaB or hbd), and tesB. / (cont.) Supplying glucose to E. coli expressing these enzymes resulted in 3HB production, while supplying glucose and propionate results in 3HV production. Supplying glucose and glycolate resulted in DHBA production with some 3-HBL, but only with the help of a fourth gene - propionyl-CoA transferase (pct). Removing the tesB gene from this four-gene system substantially increases 3-HBL titers at the expense of DHBA. This work represents the first successful production of DHBA and 3-HBL in a biological system from carbohydrate-based substrates. In each of these systems, several broadly-applicable tools and strategies were developed to enhance product titer or discover new metabolic activities. In the P. putida system, cytosolic and extracytosolic biocatalysis were combined in a single metabolic pathway to realize lactone production. This catalytic strategy, termed integrated bioprocessing, is applicable to other metabolic pathways whose production suffers due to a suboptimal cytosolic enzyme. Also in the P. putida system, two-phase cultures were used to sequester the lactone product away from the lactonase, helping to drive lactonehydroxyacid equilibrium towards the lactone. This methodology allows one to overcome equilibrium-based limitations of product titer. Finally in the E. coli work, a promiscuous pathway normally used for polyhydroxyalkanoate synthesis was exploited to give a wide range of hydroxyacid products. This substrate promiscuity was critical in achieving the production of new compounds biologically and thus substrate promiscuity was identified as a key component for metabolic pathway design and construction. / by Collin H. Martin. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Chemical Engineering, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 221-235). / Low temperature oxidation in the gas and condensed phases has been the subject of experimental investigations for many decades owing to applications in many areas of practical significance like thermal stability, combustion, atmospheric chemistry and industrial syntheses. Owing to several practical limitations it has proven difficult to understand these processes at a mechanistic level from experiments alone. Developments in scientific computing have opened up computational chemistry and cheminformatics based tools as an attractive option for exploring and elucidating the kinetics of these complex processes through detailed kinetic modeling and requires efforts in three key areas: single reaction kinetics, reaction networks and coupling kinetics with mass/momentum/energy balance models. This thesis presents several contributions employing high-level electronic structure calculations, reaction rate theory, automated kinetic modeling and empirical correlations to further our mechanistic understanding of low-temperature oxidation in the gas and liquid phase. First, an extensible framework for automatic estimation of species thermochemistry in the solution phase is presented and validated. This framework uses the Linear Solvation Energy Relationship (LSER) formalism of Abraham/Mintz and co-workers for high-throughput estimation of [delta]G°solv(T) in over 30 solvents using solute descriptors estimated from group additivity. The performance of scaled particle theory (SPT) expressions for enthalpic-entropic decomposition of [delta]G°solv(T) is also discussed along with the associated computational issues. Second, the importance of solvent effects on free-radical kinetics is explored using tetralin oxidation as a case study. The solvent dependence for the main propagation and termination reactions are determined using the Polarizable Continuum (PCM) family of solvation models. Incorporating these kinetic solvent effects in detailed kinetic models suggest oxidation rates increase with solvent polarity, consistent with experiment. Following this, electronic structure methods and reaction rate theory are used elucidate mechanistic details of new pathways in liquid-phase and atmospheric oxidation. The first of these studies focuses on pathways that establish [gamma]-ketohydroperoxides (KHP), well-known products in low-temperature alkane oxidation, as precursors to acids through a two-step process. Ab initio calculations are used to identify pathways leading from KHP to a cyclic peroxide isomer which decomposes through novel concerted reactions into carbonyl and carboxylic acid products. High-level gas phase rate coefficients are obtained using DFT/WFT methods coupled with VTST/SCT calculations and multi-structural partition functions (QMs-T). Solvent effects are included using continuum dielectric solvation models and the predicted rate coefficients found to be in excellent agreement with experiment lending theoretical support to the 30-year old Korcek hypothesis. Next, insights from the Korcek reaction are extended to atmospheric chemistry where similar cyclic peroxides are formed by reactions of the Criegee Intermediate (*CH₂OO*) with double bonds. More specifically, the role of chemical activation in reactions between *CH₂OO* and C=O/C=C species is explored using master equation calculations to obtain phenomenological rate coefficients k(T,P). In the case of reactions with C=O, the yield of collisionally stabilized SOZ at atmospheric pressure was found to increase in the order HCHO < CH₃CHO < CH₃COCH₃ - At low pressures, chemically activated formation of organic acids was found to be the major product channel in agreement with recent direct measurements. Epoxide and CH₂=CHOH are predicted to be the major products for *CH₂OO* + C₂H₄ under atmospheric conditions. Finally, as a case study in coupling detailed chemical and physical models, the improved understanding of liquid phase oxidation developed above is used to build multi-physics models of diesel injector deposit formation that adversely affects fuel spray characteristics and engine efficiency. Octane is used as a model liquid fuel for detailed kinetic modeling of oxidative aging leading to deposit precursors. In addition to fuel chemistry, the immiscibility of polar oxidation products leading to 'soft deposit' is modeled using linear solvation energy relationships. The chemistry and phase separation models are coupled with physical processes like washing. The resulting framework is used to explore the sensitivity of deposit formation to various model parameters. / by Amrit Jalan. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1998. / Includes bibliographical references (p. 163-170). / by Jane Pei-chen Chang. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Chemical Engineering, 2018. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 187-210). / Chemical and biological systems are increasingly implemented with advanced sensor systems that collect large amounts of data. For example, a single microarray can measure thousands of genes and a typical offshore oil platform generates 1 to 2 TB of data per day. New algorithms are needed to efficiently and effectively use these datasets to increase predictive capability and improve system understanding. In this thesis, algorithmic advances to bridge the gap between data and system insights are addressed in a series of case studies. In the first case study, the problem of predicting critical quality attributes for a monoclonal antibody using data from the manufacturing process is addressed. In this setting, the main challenge is that there is only a limited dataset available for modeling. To tackle this issue, Monte Carlo sampling was used in conjunction with an elastic net approach to subset selection. The second case study is also within the biological domain but considers a discrete outcome. The proposed algorithm addresses two common issues when building classification models for biological studies: learning a sparse model, where only a subset of a large number of possible predictors is used, and training in the presence of missing data. The resulting algorithm leverages expectation-maximization to tackle both issues simultaneously. In the third case study, the goal was to identify anomalous operating periods using production data from an oil and gas well without access to historical examples of such periods. The proposed approach recasts the problem as a semi-supervised problem and leverages approaches from the positive and unlabeled literature. The final case study considers the task of prediction lithium-ion battery cycle life. Cycle life is defined as the number of charge and discharge cycles the battery undergoes before 80% capacity fade. Several, difficult to identify factors can contribute to capacity fade. Even in batteries with the same chemistry, operated using the same conditions, there is considerable cycle life variability. Therefore, the challenge was to build a model to capture individual capacity trajectories. Each case study is benchmarked using state-of-the-art approaches. In all settings, the value of data-driven methods is demonstrated. / by Kristen Ann Severson. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2010. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 176-182). / The electrostatic layer-by-layer (LbL) assembly technique can be used to make uniform, conformal multi-stack nanoparticle thin films from aqueous solution, with precise thickness and roughness control over each stack. Much of the effort in this area has focused on the assembly and characterization of novel nanostructures. However, there is a scarcity of studies addressing critical barriers to commercialization of LbL technology, such as the lack of mechanical durability and the difficulty of incorporating a diverse set of functional organic molecules into aqueous solution-based nanoparticle assemblies. The versatility of existing chemical functionalization methods are limited by requirements for particular substrate surface chemistries, compatible solvents, and concerns over uncontrolled nanoparticle deposition. Here we describe the advantageous use of capillary condensation, a well-known natural phenomenon in nanoporous materials, as a more universal functionalization strategy. Capillary condensation of solvent molecules into nanoporous LbL films was shown to bridge neighboring nanoparticles via a dissolution-redeposition mechanism to impart mechanical durability to otherwise delicate films. In situ crosslinking ability of photosensitive capillary-condensates was demonstrated. Particle size-dependence of the capillary condensation process was studied theoretically and utilized experimentally to modulate refractive index over coating thickness to achieve broadband antireflection (AR) functionality. Multi-stack AR coatings with alternating high- and low-index stacks were also made, and the influence of inter-stack and surface roughness on film transparency were studied quantitatively. The equivalent-stack approximation was utilized and presented as an enabling design tool for fabricating sophisticated solution-based optical coatings. Surface wettability could also be modified using capillary condensation - either by condensation of adventitious vapors during an aging process leading to a loss of optimized film properties, or by advantageous condensation of carefully chosen hydrophobic or hydrophilic molecules to tune wettability. Finally, preliminary Young's moduli measurements of all-nanoparticle and polymer-nanoparticle composite films were made using strain induced elastic buckling instabilities for mechanical measurements (SIEBIMM). / by Gemici Zekeriyya. / Ph.D.

Thesis (Sc. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 1943. / Vita. / Includes bibliographical references (leaf [167]). / by John Ploeger Longwell. / Sc.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2005. / Includes bibliographical references. / Objectives: The aims of this investigation are to (i) prepare hybrid quantum dot (QD)-polymer compleses, (ii) maniplulate structural and chemical properties of the hybrids and characterize their effects on biocompatibility, and (iii) assemble diverse, heterostructured complexes for enhancing and fluorescently tracking gene therapy. / by Saeeda Mahdi Jaffar. / Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2009. / Includes bibliographical references (leaves 194-201). / The present work is motivated by the desire to understand the physical mechanism underlying the propagation of nervous influx in neurons as well as the modulation and summation of electrical signals during their progression in dendrites. The survey of existing literature shows that most studies model dendrites and axons as cables and simply ignore the physical basis of these electrical signals. Indeed in neurons as in any cell there exists a potential difference of about 70 mV across the plasmic membrane caused by differences in the concentrations of small ions between the inside and outside of the cell; nervous influx is a temporary alteration of this potential difference that can propagate along the plasmic membrane. Nervous influx is carried by ionic currents flowing both across the membrane and inside the neuron under the membrane. It propagates at velocities comprised between 0.1 m/s and 100 m/s. The present study ambitioned to build a model for the propagation of nervous influx based on ionic currents but the literature on electrolytes did not provide a good fundamental theory to construct such a model. Therefore the subject of the thesis evolved from modeling nervous influx to developing a new fundamental theory of ionic solutions and testing it against available experimental data. The reference theory was elaborated by Debye and Hickel in the 1930's and is the only full and consistent theory but it is valid only in dilute media because it is insufficient on two points: it considers ions as infinitesimally small and ignores their correlations. Consequently a new theory of small ions is developed in which ions are treated as hard spheres and their correlations are included locally. The new theory is consistent and can be reduced to the Debye-Hiickel theory when the radius of ions and the correlation length are taken to be zero. The theory assigns to each ion an ionic radius corresponding to the extent of its electronic cloud and a hydrated radius corresponding to the extent of its shell of hydration. The solution as a whole has a correlation length determining the volume around each ion in which correlations have a significant effect. The correlation length and the hydrated radii are determined from experimental data, whereas the ionic radii have a fixed value taken from the literature. The correlation length and the hydrated radii were determined for 75 binary electrolytes by calculating the osmotic coefficient and the mean ionic activity coefficient and by fitting the wealth of experimental data available. Usually three fits were made for each electrolyte. The general fit is always close to experimental points but underestimates the values in the semi-dilute region. The dilute fit and the semi-dilute fit approximate experimental points very well in their respective regions but markedly stray from them in the concentrated region. The main explanation for the need for three fits is that the correlation length and the hydrated radii actually change with the concentration of ions whereas the model assumes them to be constant. The rate of change being small, the whole set of experimental points can be approximated very well with three fits. et of experimental points can be approximated very well with three fits. The next step was to compare these three fits with the semi-empirical models of Meissner and Pitzer for the mean ionic activity coefficient. All three are good approximations. The Pitzer model is the most precise for almost all electrolytes; the Meissner model and the present model are equivalent in their accuracy. Afterwards a short study was performed on multicomponent electrolytes and it confirmed the behavior observed for binary electrolytes. The osmotic coefficient and the mean ionic activity coefficient are usually well approximated in the concentrated and dilute regions and slightly underestimated in the semi-dilute region. The last part of the present study examines the propagation of linear plane waves. In bulk solutions the propagation is generally conservative and the phase and group velocities are on the order of hundreds of meters per second. The group velocity in bulk solutions provides an upper bound for the velocity of nervous influx; the velocity resulting from diffusion provides a lower bound; the actual mechanism must be a combination of the two depending on local conditions. In a nutshell the principal objective has been fulfilled: a new theory for ionic solutions has been developed that is valid from infinite dilution to saturation; it reproduces experimental data at equilibrium well. The theory is then applied to describe the propagation of linear waves in bulk solutions and has found group velocities on the order of hundreds of meters per second. Nevertheless the equations studied thus far cannot be applied to describe the environment under the plasmic membrane of neurons because they assume that local concentrations are small perturbations from the bulk concentrations, which is not the case. Thus the other objective of the thesis is only partially fulfilled; ionic waves are a plausible mechanism for the propagation of nervous influx and the most probable one, but the demonstration is incomplete. / by Mihai Anton. / Ph.D.