University of Illinois Urbana-Champaign
Illinois Digital Environment for Access to Learning and Scholarship RepositoryNot a member yet
123813 research outputs found
Sort by
Gravitational operator algebras and the role of noncommutative conditional probability in quantum geometry
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Marc Klinger, accepted the attached license on 2025-06-20 at 10:22.The student, Marc Klinger, submitted this Dissertation for approval on 2025-06-20 at 10:35.This Dissertation was approved for publication on 2025-07-01 at 12:53.DSpace SAF Submission Ingestion Package generated from Vireo submission #22347 on 2025-10-20 at 16:57:12A central goal of quantum gravity is to fit geometry into the formalism of quantum mechanics. In the last several years, it has been appreciated that the role of geometry in quantum theory is in fact very multifaceted. This reflects and expands upon the dual role played by geometry in classical physics; not only is geometry a classical dynamical feature in its own right, it is also a fundamental structure on top of which every other dynamical feature is built. Likewise, the quantization of geometry does not only entail a promotion of classical geometric variables to quantum operators, but also a rather intimate realignment of the fundamental structure of the full algebra of observables of any system in which dynamical, quantum gravity is present. In this thesis, we undertake a detailed study of the multifaceted role of geometry by observing how it is encoded in classical, semiclassical, and fully quantum analyses of gravity. In Chapter 2, we introduce the extended phase space as a symplectic geometric approach to accounting for the full set of dynamical degrees of freedom in subregions for general gauge theories and gravity. The extended phase space underscores the important role played by large gauge transformations, supported on codimension two submanifolds of spacetime called corners, which are genuine symmetries of a gauge theory rather than degeneracies of its symplectic form. The inclusion of Noether charges generating large gauge transformations as Hamiltonian functions is a defining feature of the extended phase space. In Chapter 3, we propose a quantization of the extended phase space in terms of a von Neumann algebraic construction called the crossed product. The crossed product quantizes both the ordinary phase space and the aforementioned Noether charges to operators acting on an extended Hilbert space. In the gravitational context, the inclusion of these extended degrees of freedom has profound implications for the entanglement structure of the resulting theory. Familiar divergences which are encountered in ordinary quantum field theories are lifted and rigorous notions of density operators and von Neumann entropies can be assigned to states in the resulting subregion algebras. In Chapter 4, we formalize this result by proving a theorem which establishes when the crossed product of a type III von Neumann algebra with a locally compact group is semifinite. We show that this will be the case provided admits as a subgroup the modular automorphism of a faithful, semifinite, normal weight on , and is (quasi)-invariant with respect to the action of on . Gauge invariant measurements are intrinsically conditional -- depending upon the specification of a dynamical quantum reference frame which plays the role of a measuring apparatus. In Chapter 5, we identify the new degrees of freedom central to the extended phase space and the crossed product as quantum reference frames. These degrees of freedom allow for fields and operators in the non-extended theory to be dressed to satisfy constraints associated with genuine gauge symmetries. Crucially, these reference frames are formed from degrees of freedom from within the theory; quantum gauge theories come equipped with the ability to measure themselves. This observation resounds the important role of large gauge transformations in facilitating the gluing of subregions in a manifestly gauge invariant fashion. In Chapter 6, we contextualize the preceding results by observing that the extended phase space and the crossed product can be understood as subcases of a more general object called a quantum orbifold. Finally, in Chapter7, we analyze the quantum information theoretic properties of gravitational subregion algebras in light of the extensions addressed above. We demonstrate that the von Neumann entropy of a generic state in the gravitational crossed product algebra can be interpreted directly as a generalized entropy in the holographic sense. Moreover, we illustrate that this generalized entropy can be regarded as a factorized entropy under the presence of a (noncommutative) conditional expectation implementing a form of non-exact quantum error correction. This observation motivates a definition for the area operator in fully non-perturbative quantum gravity as the log of a relative conditional density operator whose expectation value computes the conditional entropy of a chosen state
Design and demonstration of a motional narrowing method to measure a low temperature 3He diffusion coefficient for a superfluid neutron electric dipole moment experiment
No full textSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Cameron Erickson, accepted the attached license on 2025-07-07 at 08:29.The student, Cameron Erickson, submitted this Dissertation for approval on 2025-07-07 at 09:23.This Dissertation was approved for publication on 2025-07-09 at 13:30.DSpace SAF Submission Ingestion Package generated from Vireo submission #22420 on 2025-10-20 at 16:57:28Measurements of the neutron electric dipole moment (nEDM) are cosmologically motivated by a search for CP symmetry violations in particle physics. Despite over six decades of effort, physicists have yet to attain the precision necessary to detect a nonzero nEDM. An experiment technique first proposed in \cite{GolubLamoreaux}, is expected to be capable of achieving a precision 2 orders of magnitude smaller than the current leading bound \cite{SNSnEDM}. The key distinction of this technique is that the neutrons are measured in a superfluid 4He bath through spin dependent interactions with 3He – warranting the "superfluid nEDM experiment" label. Important to the design of a superfluid nEDM experiment is a knowledge of the transport properties of 3He in the superfluid 4He. This information is partially encoded by a "3He-phonon" diffusion coefficient denoted by which describes the diffusion of 3He against the superfluid 4He phonon excitations in the liquid. All previous measurements of used local 3He density dependences on temperature gradients to extract . The most recent measurement by Rao \cite{Rao} obtained a result that disagrees with theory \cite{BBBVeryDilute},\cite{BBBDilute} by a factor of 3. This thesis proposes, designs, and demonstrates a different method to measure that does not use temperature gradients, but instead centers around use of the "motional narrowing limit" of nuclear magnetic resonance free induction decays. To the author's knowledge, this motional narrowing technique has been used only once before to determine unknown diffusion constants by \cite{Himbert}. An important difference of the work here from \cite{Himbert} is that the 3He is spin polarized at room temperature and then injected into cryogenic measurement cell as opposed to being polarized directly in a cryogenic measurement cell. Part of the contribution of this thesis is a discussion of the design of such an experiment to measure . It is argued that accessing the experimental parameter space required by the narrowing method is not obvious. This motivates measurements at about 45 Kelvin using only 3He gas to demonstrate 3 key attainable design elements: the "injection efficiency", minimum background magnetic gradients, and statistical precision of the diffusion coefficient extraction The injection efficiency corresponds to the fraction of remaining spin polarization after transport between room temperature to 45 K. The measurements in this thesis demonstrate an injection efficiency of up to . Minimal background magnetic gradients are achieved with a 5 coil shimming system (the "pentacoils") and demonstrated to be equivalent to Gauss/cm over the span of a few centimeters. Diffusion coefficients measurements span from 46 to 372 cm/sec are determined with a statistical precision as low as for applied gradients equivalent to . The systematic uncertainties are much larger, but these could be significantly reduced by future works
Detailed analysis of ligands on gold nanoparticles
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Katherine Hatzis, accepted the attached license on 2025-07-09 at 13:26.The student, Katherine Hatzis, submitted this Dissertation for approval on 2025-07-09 at 13:35.This Dissertation was approved for publication on 2025-07-10 at 14:31.DSpace SAF Submission Ingestion Package generated from Vireo submission #22462 on 2025-10-20 at 16:57:34Gold nanoparticles are utilized in several fields including sensing, biological therapies, and catalysis due to their relatively inert surfaces, tunable optical properties, and light-to-heat conversion. The surfaces of these nanoparticles are frequently coated with organic ligands to impart colloidal stability and engineer functionality, serving as the face of the nanoparticle to the environment. This dissertation seeks to further our current understanding of how organic ligands bind to, arrange on, and protect gold nanoparticle surfaces. In Chapter 1, the current state of the literature on gold nanocrystal surface chemistry is reviewed. Background on the properties of gold nanoparticles and how surface chemistry affects their eventual application is mentioned. Finally, some recent literature examples of ligand environment analysis on gold nanoparticles using state-of-the-art methods are presented and discussed. The work described in Chapter 2 presents a systematic investigation into how ligand length and size of gold nanoparticles affect ligand structure, dynamics, and mobility using 1H solution NMR, cyanide etching experiments, and molecular dynamics simulations. A library of gold nanoparticles with 4 average diameters (2, 4, 9, and 12 nm) and appended with 4 different mercapto-(X-alkyl)-N,N,N-trimethylammonium bromide (MxTAB) ligands (X= 11, 16, 18, 20) was synthesized, and patterns in NMR peak shift, ligand density quantification, and T2 relaxation constants were used to draw conclusions about the bifactorial effects of gold nanoparticle diameter and ligand length. Additionally, cyanide etching experiments and molecular dynamics simulations gave additional clues to the overall structure and character of the ligand shell on each gold nanoparticle type. In Chapter 3, the impact of the synthesis and functionalization method, and the intermediates used therein, on the resulting character of the ligand shell are investigated. Nominally similar 5 nm diameter gold nanospheres coated with identical cationic ligands were synthesized and functionalized through 3 methods, their ligand shells characterized, and the nanoparticles’ resistance to cyanide etching and morphology change with thermal annealing were investigated. Future experiments were proposed to further investigate how intermediates used in processing of the nanoparticles could affect the character of the ligand shell. Finally, in Chapter 4, N-heterocyclic carbenes (NHCs) were investigated through density functional theory and spectroscopy methods to determine the kinetics and thermodynamics of ligand binding and etching to gold nanoparticle surfaces. We discovered a negative correlation between gold-NHC binding energy and extent of gold nanoparticle etching, which is opposite of conventional thinking about surface etching. We proposed a ligand pKa-related model to corroborate these surprising findings
Bottom-up theoretical frameworks for upscaling transient mass transport in porous media
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Md Abdul Hamid, accepted the attached license on 2025-07-10 at 03:47.The student, Md Abdul Hamid, submitted this Dissertation for approval on 2025-07-10 at 03:50.This Dissertation was approved for publication on 2025-07-13 at 10:54.DSpace SAF Submission Ingestion Package generated from Vireo submission #22479 on 2025-10-20 at 16:57:41Solute transport in porous media plays a fundamental role in a wide range of engineering, environmental, and biomedical systems. Understanding how dissolved species move through these complex structures, where advection, diffusion, and interfacial reactions all contribute, is essential for optimizing the performance of electrochemical reactors, groundwater remediation technologies, and biological tissue scaffolds. However, the multiscale nature of these transport phenomena presents significant modeling challenges. Conventional upscaling approaches often rely on empirical correlations and time-invariant effective parameters, which overlook critical pore-scale dynamics, particularly under transient conditions. This thesis addresses these shortcomings by developing bottom-up theoretical frameworks that systematically link pore-scale transport physics with macroscopic behavior under time-varying conditions. The first framework begins with a pseudo-steady approximation suitable for saturated porous media experiencing slow transient relative to pore-scale diffusion. This formulation is developed using a representative unit cell under the assumption of negligible macroscopic diffusion (i.e., dispersion). Its validity is assessed using nondimensional parameters (Péclet, Sherwood, and Fourier numbers) through dynamic analysis of redox flow battery cycles. The reactive liquid/solid interfaces are modeled using a spatially uniform, time-dependent Dirichlet boundary condition, justified under the assumption of a large Damköhler number (Da≫1). Additional validity criteria are established for three classes of electrochemical reactions, accounting for variations in kinetics and stoichiometry. Building on this foundation, a second framework adopts a spectral bottom-up approach to study porous systems where transient effects are significant, but macroscopic concentration gradients remain negligible. The mass conservation equations are transformed into the frequency domain using Fourier transformation, resulting in two frequency-dependent, volume-averaged transport coefficients interpreted as transfer functions (TFs). The first, the spectral Sherwood number, quantifies interfacial reactive flux. The second TF captures deviations from ideal advection due to microscale variations in local velocity and concentration fields. These transfer functions are integrated into a bottom-up transient model (B-UTM) of a redox flow cell to simulate dynamic polarization behavior. The B-UTM exhibits improved agreement with experimental data and outperforms conventional models that rely on a constant Sherwood number. Additionally, it enables simulation of near- and over-limiting current operation, which traditional models cannot capture. A regime map generated from the model illustrates that short-duration over-limiting operation is feasible, suggesting a strategy to enhance charge capacity in redox flow battery systems under transient loading. To overcome the limitation of assuming negligible macroscopic gradients, a third theoretical framework, the omni-temporal theory, is introduced. This framework extends the frequency-domain model by incorporating macroscopic diffusion (i.e., dispersion) through rigorous volume averaging. A key element of the theory is an ansatz that assumes a linear relationship between local concentration deviations and the macroscopic concentration gradient. This leads to two closure problems, the solutions of which enable systematic upscaling. Three frequency-dependent, Darcy-scale transport coefficients are derived and interpreted as transfer functions: an effective reaction rate TF, an advection suppression TF, and an effective diffusion tensor. The first two extend previous transfer functions to systems with non-negligible macroscopic gradients. The diffusion tensor captures the combined effects of molecular diffusion, dispersion, and tortuosity. Notably, the framework reveals that tortuosity is not solely a geometric property but emerges dynamically from microscale transport interactions that depend on frequency. The omni-temporal theory is validated by numerically solving the frequency-dependent advection–dispersion–reaction equation and benchmarking its predictions against microscopically resolved direct numerical simulations (DNS). The model demonstrates excellent agreement with DNS and outperforms traditional models that use time-invariant diffusion coefficient. It accurately captures transient dispersion in both pre-asymptotic and asymptotic regimes, offering a unified multiscale framework for modeling time-dependent transport in porous media. The modeling approaches developed in this thesis evolved from pseudo-steady to spectral and finally to omni-temporal formulations, each providing increased accuracy and broader applicability across a range of transient regimes. These frameworks enable reliable simulation of time-varying transport processes and are applicable to systems including electrochemical devices, CO₂ sequestration, enhanced oil recovery, drug delivery, and contaminant migration in aquifers. Furthermore, the theoretical frameworks developed in this study can be extended to heat transfer problems by leveraging the mathematical analogies between heat and mass transport
Hybrid decomposition-based control co-design of energy systems using graph-based models
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Kayla Smith, accepted the attached license on 2025-07-10 at 19:11.The student, Kayla Smith, submitted this Dissertation for approval on 2025-07-10 at 19:13.This Dissertation was approved for publication on 2025-07-11 at 15:50.DSpace SAF Submission Ingestion Package generated from Vireo submission #22487 on 2025-10-20 at 16:57:43Energy systems, defined in this work as systems that generate, use, transfer, or store energy, are critical to everyday life. Designing energy systems to be efficient is an important task, because energy system performance has significant environmental and economic impacts. Within energy system design, there are two design questions that can be asked. The first is “What is the best overall design for the system?”. The second is “What is the best subsystem to optimize and how do I optimize it?”. This dissertation seeks to answer both types of design problems. There are several challenges that arise with energy system design. First, many energy systems are dynamic systems that require controllers to actuate the system to achieve desirable performance. To adequately design the system, both the plant and control design need to be considered together. Second, with technological advances, many energy systems are more interconnected than in the past, resulting in large systems consisting of several subsystems each with their own performance goals. Energy system design frameworks must be capable of optimizing large interconnected systems. Lastly, many energy systems are designed using existing components from manufacturers, resulting in a component selection problem. Energy system design algorithms should incorporate discrete component-based design. To address these challenges, this dissertation proposes a hybrid, decomposition-based control co-design approach. This framework combines two different design frameworks to address all three challenges. The hybrid optimization framework addresses the component-based design problem by converting the discrete component selection problem into a continuous problem. Then after the continuous solution is found, a sorted search algorithm is applied to efficiently find the optimal combination of components. The second framework is a decomposition-based control co-design framework. This framework partitions the system based on the dynamics of the system, resulting in subsystems that have little power flow between them. Then the subsystem problems are coordinated and solved to ensure consistency among the solutions. This framework is applied to two different case studies: a thermal management system and a quadrotor system. In the thermal management system case study the decomposition-based control co-design portion of the framework is applied and studied. This case study demonstrates the ability of the decomposition-based framework to converge for a single energy domain system. In the quadrotor case study, the full hybrid, decomposition-based approach is applied. This case study demonstrates the computational benefits of the full framework and is able to solve the component selection problem. Additionally, this case study demonstrates the effectiveness of the proposed methods on a multi-domain electro-mechanical system. Overall, the two case studies demonstrate the ability of the proposed framework to efficiently optimize systems while addressing the component selection problem
Hearing Like A City: Sonic Gentrification and the Nighttime Economy of Urban Soundscapes
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Colt Pierce, accepted the attached license on 2025-07-11 at 15:12.The student, Colt Pierce, submitted this Dissertation for approval on 2025-07-11 at 15:22.This Dissertation was approved for publication on 2025-07-14 at 14:02.DSpace SAF Submission Ingestion Package generated from Vireo submission #22522 on 2025-10-20 at 16:57:52Austin, Texas, has emerged as an exciting city to study among urbanists Now regarded as a premier technopolis and cultural center in the Sun Belt, the city has undergone rapid change and dramatic gentrification, with many questioning whether its previous cultural character, lauded by Bohemian expatriates, remains intact. Studies of Austin have examined numerous things, particularly its ongoing gentrification, which has proved complicated. This research builds upon this emergent body of gentrification research in Austin to chronicle that a glaring omission marks this work: the driving influence of sound in facilitating this urban restructuring. The literature on gentrification primarily focuses on the visual: ocular-centric takes on demographic changes, housing upgrades, and budding boutiques. While some scholars have examined the politics of sound and gentrification, little attention has been given to understanding how cities control sound as an active ingredient that induces neighborhood change and valorizes land. This research introduces the concepts of sonic gentrification and sonic fix. Focusing on music-oriented development in East Austin, this research reveals how sound and live music are injected into the logic of urban growth to purposefully change the sounds of a neighborhood, promote a middle class takeover, push out Latine and Black residents, and close rent gaps as it creates its nighttime economy. Using archival work, sound walks, and ethnographic methods, this research codifies this process by illuminating the racialization, erasure, and commodification of sounds, demonstrating that city soundscapes are manipulated and embedded within systems of power. In this frame, my work urges geographers to hear the city as a powerful dimension promoting urban change
Natural language processing for supporting impact assessment of funded projects
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Kanyao Han, accepted the attached license on 2025-07-14 at 23:45.The student, Kanyao Han, submitted this Dissertation for approval on 2025-07-14 at 23:53.This Dissertation was approved for publication on 2025-07-15 at 16:19.DSpace SAF Submission Ingestion Package generated from Vireo submission #22526 on 2025-10-20 at 16:57:52Funding from organizations like the U.S. National Science Foundation plays a crucial role in supporting researchers and practitioners in advancing scientific knowledge, promoting societal progress, and protecting the environment, among other goals. As a result, both organizations and researchers are keen to understand how such funding is distributed across various projects and disciplines, as well as the outcomes and impacts generated by these projects. A comprehensive analysis of diverse text-based data sources that document funding allocations, research outcomes, and broader impacts can help deepen this understanding. These data sources include project reports submitted to funders as well as outcomes published in research articles. However, annotating and analyzing text-based data, even at moderate volumes, can be time-consuming and costly. Researchers must process lengthy and large-scale datasets to identify meaningful information for analysis. This dissertation aims to leverage computational methods, particularly from the fields of Natural Language Processing (NLP) and Machine Learning (ML), to assist researchers and practitioners in managing text-based data more efficiently and effectively. By automating or semi-automating processes such as information extraction, data cleaning, and classification, this work seeks to reduce the workload associated with data processing and annotation. This dissertation explores how NLP and ML techniques can be developed and used to handle data from social and scientific research under three challenging conditions: (1) disorganized, complex, lengthy, or incomplete datasets; (2) limited availability of annotated data; and (3) the need for domain-specific analysis schemas. By addressing these challenges, this dissertation aims to develop innovative approaches to aid in the analysis of funding allocation and the assessment of the impact of funded projects, with three studies being presented. First, analyzing past funding allocations can offer valuable insights into funding patterns in previous research. However, such analyses are often hindered by inconsistent and ambiguous naming conventions for funding organizations in publication records. This dissertation proposes a framework for fine-tuning a model to disambiguate funder names. Second, categorizing project reports can provide valuable insights into how funding is allocated across different project themes. Despite the availability of various categorization methods that typically require large volumes of annotated data for model fine-tuning or training, little is known about how to build effective models when: (a) categorizing texts demands substantial domain expertise and/or detailed reading; (b) only a limited number of annotated documents are available for training; and (c) no relevant computational resources, such as effective pre-trained models, exist. This dissertation introduces and evaluates a categorization method that combines expert knowledge with computational models to develop domain-specific categorization models. Third, with funding agencies increasingly demanding evidence of the social impact of scientific research, impact assessment has become critical. However, challenges remain in categorizing research reports due to the absence of a comprehensive impact classification schema and standardized reporting formats across domains. This dissertation addresses these gaps by developing and evaluating a classification schema for assessing the impact of funded research projects across domains, assisted by NLP and ML techniques. This dissertation advances knowledge by (1) developing novel frameworks for cleaning, annotating, and extracting valuable information from publication records and project reports; (2) providing insights into funding allocation in scientific research and biodiversity conservation; and (3) enhancing the understanding of the impacts described by funded projects
Development of diagnostic, therapeutic, and multi-purpose agents for Alzheimer's disease
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Karna Terpstra, accepted the attached license on 2025-07-15 at 16:10.The student, Karna Terpstra, submitted this Dissertation for approval on 2025-07-15 at 16:20.This Dissertation was approved for publication on 2025-07-16 at 13:04.DSpace SAF Submission Ingestion Package generated from Vireo submission #22586 on 2025-10-20 at 16:58:34Alzheimer’s disease is a progressive neurodegenerative disease that, despite discovery in the early 20th century, remains incurable. Additionally, the pathology of the disease remains unknown with many neurological abnormalities including amyloid-β plaques, phosphorylated tau neurofibrillary tangles, neuronal signaling abnormalities, increased metal ion concentrations, reactive oxygen species, and neuroinflammation all having been indicated in the progression of the disease. Among the earliest and most characteristic biomarkers of Alzheimer’s disease is the presence of amyloid-β peptide aggregates, making them an area of great interest for diagnostic agent development. Positron emission tomography agents incorporating 64Cu have been of great interest due to the long half and relatively simple incorporation by chelation of the radionuclide. In the second chapter of this work, a series of amyloid-β targeting, 64Cu chelating imaging agents were synthesized. Analysis of the complexes concluded that the presence of phenolates and carboxylic acids on the ligand increased the complex stability, but the presence of alkyl substituents close to the chelation site of the radionuclide detrimentally impacted both the stability and the lipophilicity of the complexes. This indicates that the addition of alkyl substituents close to the site of radionuclide chelation is not a beneficial tactic in increasing the lipophilicity and therefore the brain uptake of the complexes. Agents that impact the aggregation pathway of amyloid-β are of great interest for therapeutic agents and for the investigation of Alzheimer’s disease pathology. Iridium complexes have great potential in the alteration of amyloid-β aggregation pathway due to their biocompatibility, optical properties, and modular synthesis. In the third chapter of this work, a series of amyloid-β interacting Ir(III) complexes were synthesized and evaluated for their ability to enter the brain and impact the aggregation pathway of amyloid-β. HN-1, a complex with two open coordination sites was seen to significantly prevent the formation of amyloid-β plaques, likely due to the interaction of Ir(III) with amyloid-β residues. While quantification of Ir(III) complex brain uptake was hindered by solubility and likely limited blood brain barrier permeability, several complexes were shown to be present in the brains of 5xFAD mice and interact with amyloid-β. Therefore, Ir(III) complexes are promising agents for altering the aggregation pathway of amyloid-β. Several small molecule therapies are currently utilized in the clinic as Alzheimer’s disease treatments. However, the efficacy of these therapies can decrease over time and result in the side effect incidence outweighing the benefits of the therapy. Targeting donepezil and memantine therapies to the area of amyloid-β plaques may increase the efficacy of the therapies and decrease the prevalence of side effects. In the fourth chapter of this work, novel derivatives of memantine and donepezil were synthesized and evaluated as dual-function compounds. The donepezil derivatives were observed to have low inhibition capacity, indicating the mechanism of multifunctionality detrimentally impacted the interaction of the donepezil with acetylcholinesterase. However, the memantine derivative was observed to have prodrug activity, breaking down to release memantine in the brains of 5xFAD mice. Therefore, the development of donepezil derivatives by this mechanism was not beneficial, but the continued optimization of a dual-function memantine prodrug may elongate the treatment efficacy of memantine. Overall, the development and thorough understanding of agents to alter the aggregation pathway, image, and target treatments to the protein contribute to the understanding of the role amyloid-β aggregates in Alzheimer’s disease
Species transfer processes in molten salt reactors
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Joon Hon Alvin Lee, accepted the attached license on 2025-07-17 at 11:18.The student, Joon Hon Alvin Lee, submitted this Dissertation for approval on 2025-07-17 at 12:31.This Dissertation was approved for publication on 2025-07-17 at 16:41.DSpace SAF Submission Ingestion Package generated from Vireo submission #22634 on 2025-10-20 at 16:58:43In a molten salt reactor (MSR), the accurate simulation of the fuel salt and salt-facing component compositions is especially important because the nuclear fuel is dissolved within the molten salt and a change in composition can have significant impacts on the neutronics and thermal-hydraulics behavior of the reactor. In this dissertation, a method for modeling the continuous evolution of material compositions in MSRs under various species transfer processes was developed and demonstrated. The method uses the reprocessor function within Serpent 2, which is based on a modified Bateman equation for fuel depletion and is easily extendable to other depletion codes with user-defined species (or material) transfer capabilities. The dissertation focus is on the application of the species transfer method in a graphite moderated MSR to (a) determine the target reprocessing rates of fission products, (b) determine the acceptable graphite adsorption rates and equilibrium constants based on final waste classification, (c) determine the acceptable deposition and corrosion rates of heat exchanger materials, and (d) develop and demonstrate a method for detecting plutonium diversion considering the major and minor species transfer processes in (a), (b), and (c). The results in (a), (b), and (c) can then be used by reactor designers and manufacturers as reference levels to be satisfied in order to fulfil the design requirements, while the results in (d) can be used by safeguards inspectors for a non-proliferation monitoring program. These efforts are aimed to contribute to the successful deployment of MSRs. Relations were derived for converting the physical parameters and measurable quantities of the MSR into a reprocessing rate that is required by depletion codes to simulate continuous reprocessing. Using these relations, the neutronically important elements were identified to consist of xenon and lanthanides, and their reprocessing rates leading to substantial improvements in excess reactivity were determined. Importantly, the results demonstrated that depletion calculations with continuous reprocessing are necessary in order to obtain the correct keff due to contributions from precursors of neutron absorbing species. A refueling scheme with charge balance approximation was also developed and demonstrated to account for the changing salt charge during fuel depletion and the co removal of F- ions, with the approximation producing a notable correction to the reprocessed mass. The relations to simulate graphite adsorption and salt penetration were derived and demonstrated in this dissertation, by considering the rates of adsorption and desorption to be proportional to the species concentration within the materials in contact. Using these relations in the simulation of various adsorption scenarios, the equilibrium constants leading to greater-than-Class-A radioactive waste and their neutronics impact were determined. Simulations were also performed to determine the neutronics and wall-thickness impacts of the intermediate heat exchanger due to deposition and corrosion, and to demonstrate the conversion of postulated heat exchanger and reactor performance criteria into deposition and corrosion bounding rates. A mathematical basis was derived to translate the activities of select gamma emitting fission products into the fissile isotope ratios (or fission rate ratios). Using this set of translation relations, a methodology for the detection of plutonium diversion in MSRs was developed. The methodology consists of three key steps: i) determining the fission product yields due a single fissile isotope (e.g., 235U or 239Pu in this work), ii) determining the fuel evolution of a reference reactor with declared power history and reprocessing and refueling processes, and iii) determining the presence of plutonium diversion by comparing the indicators predicted from measurements of an evaluated reactor with the calculations of the reference reactor. The first two steps calibrated the methodology parameters and identified several pairs of gamma-emitting species that are appropriate for diversion detection due to their good predictive performance and resistance to unwanted species transfer influences. In particular, the 138mCs/134mI pair emerged as the ideal candidate due to its excellent accuracy under the extreme scenario of continuous and complete removal of its precursor elements. The performance of the detection methodology was evaluated through several plutonium diversion scenarios, and the methodology was able to accurately predict the 239Pu content in the reactor and determine the amount of diverted 239Pu in each scenario. Using the prediction uncertainty of the 138mCs/134mI pair, the sensitivity of the detection methodology was estimated to be 860 g of plutonium diversion before detection (out of 89.5 kg of plutonium generated after 364 days in a scaled-up 1000 MWth reactor). Meanwhile, using alternative candidate pairs, such as 85mKr/135mXe which require additional considerations of interfering species transfer processes, the predictive performance of the methodology can be improved significantly to 30 g of plutonium diversion before detection. This strengthening of the detection methodology performance is enabled by the accurate accounting of the influence from concurrent species transfer processes, which is addressed by the modeling methods developed in this dissertation. The new and significant contributions made in this dissertation to the body of knowledge are: -the unification of phenomena with disparate time-scales, length-scales, and physical mechanisms under a single framework that is analogous to isotope depletion (i.e., λ_transfer×N_transfer), which is trivial to implement in depletion codes supporting user-defined material transfer, -the derivation of relations to convert the physical reactor parameters and measurable physical quantities into the transfer rate constants λ_transfer, -the demonstration of utility of such approach for bounding the reactor core, material, and component performance metrics, and -the use of the above framework to develop a novel method for detecting plutonium diversion
Design and implementation of learning-based storage systems: a holistic approach
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Jinghan Sun, accepted the attached license on 2025-07-09 at 16:17.The student, Jinghan Sun, submitted this Dissertation for approval on 2025-07-09 at 16:27.This Dissertation was approved for publication on 2025-07-13 at 07:51.DSpace SAF Submission Ingestion Package generated from Vireo submission #22469 on 2025-10-20 at 20:15:08Storage systems have evolved over decades into a complicated ecosystem that includes storage hardware, system software, storage infrastructure, and data applications. This growing complexity poses significant challenges for storage development and deployment. Traditional human-driven, heuristic-based approaches to building storage systems cannot rapidly meet the ever-increasing demands for storage performance and efficiency. As we embrace the advancement of machine learning (ML), it is the golden age today to invent new approaches to building storage systems. This dissertation focuses on the design and implementation of learning-based storage systems across the entire storage stack, which develops learning-based approaches to optimize storage performance, resource efficiency, and management. The storage performance of traditional hardware and software cannot keep up with the increasing demands of applications. For example, in solid-state drives (SSDs), the flash translation layer (FTL) manages performance-critical metadata structures using human-driven heuristics, but it fails to adapt to different workload patterns and results in severe performance loss. To address this performance challenge, this dissertation proposes the first learning-based flash translation layer, LeaFTL, that can dynamically capture data access patterns of storage workloads at runtime. It significantly reduces the memory footprint of the address mapping table by grouping a large set of mapping entries into a learned segment. The saved memory space further benefits data caching and improves the overall storage performance. Storage resource efficiency is also critical as it directly affects the operational costs of storage infrastructure. For instance, cloud platforms manage storage resources at scale and it is beneficial for them to achieve high storage utilization. However, our study reveals that storage resources in modern cloud platforms are severely underutilized. To address this resource efficiency challenge, this dissertation proposes a learning-based storage harvesting framework named BlockFlex, that can dynamically harvest idle storage capacity and bandwidth to improve cloud storage utilization. BlockFlex leverages a lightweight online learning approach to predict resource utilization and storage resource demands, based on which it enables accurate storage resource harvesting. Storage management is becoming complicated with the rapid development of the software and hardware over the past decades. This is especially true for cloud platforms, where the cloud storage resource is shared by multiple tenants with complex storage states and dynamic workload characteristics. Cloud platform enforces isolation mechanisms as it manages collocated tenants, but weak isolation incurs high performance interference and strong isolation causes low storage utilization. This dissertation explores reinforcement learning (RL) techniques to combat this fundamental tussle, leveraging its unique advantages of optimizing decisions in the complex and dynamic environment. It employs multi-agent reinforcement learning to dynamically manage resource scheduling across multiple tenants. Our experiments show that it can achieve both high storage utilization and performance isolation