UWSpace

UWSpace is the University of Waterloo’s institutional repository for the free, secure, and long-term home of research produced by faculty, students, and staff.

Depositing Theses/Dissertations or Research to UWSpace

Are you a Graduate Student depositing your thesis to UWSpace? See our Thesis Deposit Help and UWSpace Thesis FAQ pages to learn more.

Are you a Faculty or Staff member depositing research to UWSpace? See our Waterloo Research Deposit Help and Self-Archiving pages to learn more.

Photo by Waterloo staff

Communities in UWSpace

Select a community to browse its collections.

Now showing 1 - 13 of 13

Arts (Faculty of)
Conrad Grebel University College
Engineering (Faculty of)
Environment (Faculty of)
Health (Faculty of)
Mathematics (Faculty of)
Renison University College
Science (Faculty of)
St. Jerome's University
St. Paul's University College
Stratford School of Interaction Design and Business
University of Waterloo
The University of Waterloo institution-wide UWSpace community.
Waterloo Library

Recent Submissions

The Function and Inhibition of SARS-CoV-2 Main Protease (Mpro)
(University of Waterloo, 2026-10-05) Mohammed, Fatma Hassan
The SARS-CoV-2 main protease (Mpro) is essential for viral replication and has no close human counterpart, making it a prime antiviral target. Yet the chemical space of covalent Mpro inhibitors remains narrow, and how these molecules recognize the active site, react with the catalytic cysteine, and are affected by the enzyme's other reactive residues is incompletely understood. This thesis addresses these questions across inhibitor discovery, enzyme engineering, and computational design. Cyclobutanone, a ring-strained electrophile inspired by penicillin's mechanism, is first shown to be a viable non-peptidic warhead for Mpro: a synthesized series inhibits the enzyme in a time-dependent manner consistent with covalent engagement, with potency governed by substituent occupancy of the hydrophobic S2–S4 subsites rather than by ring saturation. To resolve whether inhibitor potency is masked by off-target reactivity, an Mpro variant lacking its three solvent-exposed, non-catalytic cysteines is engineered. This mutant retains the wild-type fold and active-site geometry by crystallography, yet its apparent sensitivity to a covalent reference inhibitor change markedly, revealing that surface cysteines act as competing decoys in the native enzyme. The catalytic dyad itself is then dissected, with mutagenesis and crystallography showing that its two residues are functionally distinguishable: disabling the general base leaves substrate-groove geometry and covalent inhibitor trapping intact, whereas disabling the nucleophile is accompanied by localized structural rearrangement, indicating asymmetric roles in inhibitor recognition. Finally, applying molecular docking, dynamics, and free-energy calculations to a new aldehyde-based series shows that docking score and single-point binding energy are poor predictors of true potency: two computationally top-ranked compounds are among the weakest experimentally, an inconsistency resolved only by tracking pose stability across simulated time, which reveals their failure to remain productively bound. Together, these findings identify two new covalent Mpro inhibitor chemotypes, provide validated tools for isolating genuine catalytic-site reactivity, and establish binding-pose persistence, not static computational scores, as essential for reliable inhibitor triage.
Harvesting entanglement on the lattice and in holography
(University of Waterloo, 2026-10-02) Wurtz, Kelly
Entanglement in quantum field theory is difficult not only to calculate but also to interpret: continuum entanglement entropies are ultraviolet divergent, and holographic entanglement measures do not by themselves specify what a local experiment can access. This thesis studies field entanglement through localised probes: finite quantum systems coupled to quantum fields whose final states contain accessible correlations. The first part studies a free scalar field on a harmonic lattice. For a given spatial region of the field, the reduced vacuum decomposes into Williamson normal modes that isolate the entanglement between the subregion and its complement. The most entangled modes are shown to be optimal targets for extraction into detector degrees of freedom, and detector couplings are constructed that realise this extraction by swapping selected field modes with probes. These modes are supported near the boundary, so the probe degrees of freedom needed to reproduce the area-law contribution scale with boundary area rather than volume. The extraction is then assigned an energetic cost by putting selected modes in their vacuum states and evaluating the change in the original lattice Hamiltonian. Numerically, both the extracted entropy and the energy cost obey area-law behaviour over the regimes studied. The second part turns to conformal field theories and their AdS holographic duals. First, we couple Unruh-Dewitt detectors to scalar primary operators in a conformal field theory, and study their ability to harvest entanglement. At leading perturbative order, the detector state is fixed by the universal CFT two-point function, allowing harvested negativity and mutual information to be studied as functions of scaling dimension and detector geometry. In holographic CFTs, the bulk dual also provides an operational way to separate genuine harvesting from causal communication. The final part uses HKLL reconstruction to describe bulk detector interactions in anti-de Sitter space from the boundary perspective. A local bulk detector is represented on the boundary by a nonlocal, time-dependent coupling whose smearing is fixed by the reconstruction kernel.
Impacts of a Porous Surface on Flame Acceleration
(University of Waterloo, 2026-10-02) Kresnyak, Ariyanna
Deflagration-to-detonation transition (DDT) is a critical phenomenon in the development of detonation-based propulsion systems and in industrial gas safety. Recent experimental investigations by Jiang et al. and Lin et al. demonstrated that porous copper foam integrated into the walls of a square combustion channel reduced the detonation onset distance of a stoichiometric ethylene-oxygen mixture. Understanding the mechanisms responsible for this behaviour remains incomplete. This work investigates the influence of a porous boundary on flame acceleration and pressure-field development through direct numerical simulation using the compressible reacting-flow solver PeleC. Simulations were performed for a stoichiometric ethylene-oxygen mixture within a 30 mm × 30 mm square tube. The porous medium was represented using a simplified custom boundary condition consisting of evenly spaced circular outlet pores. This representation was intentionally designed to isolate surface-level mass and flow interactions while excluding combustion and transport within the internal pore structure. Post-processing techniques were developed to quantify flame propagation, shock formation, pressure-field evolution, and transport through the porous boundary. Equal-resolution comparison of the smooth-wall and porous simulations demonstrated no measurable reduction in bulk flame velocity over the available simulation interval. However, the porous boundary substantially altered pressure-wave development. Shock formation was delayed by approximately 56 μs relative to the smooth-wall simulation while occurring at a similar axial location, indicating that the porous boundary primarily influenced the rate of pressure-wave buildup. Interaction between the developing pressure field and porous boundary began prior to direct flame interaction with the pores, while mass and kinetic-energy transport increased substantially following arrival of the flame and high-pressure products. Velocity-field analysis additionally identified a reversal in transverse near-wall flow that provides a potential coupling mechanism between the bulk flame and processes occurring within the porous structure. The results demonstrate that, under the simplifying assumption of representing the porous medium as surface-level outlet pores, the experimentally observed enhancement in flame acceleration is not reproduced. The simulations instead demonstrate pressure-wave attenuation and delayed shock development while retaining similar bulk flame propagation at equal resolution. These findings indicate that physical processes excluded by the simplified boundary representation, particularly combustion and flow transport within the porous volume, may be important to reproducing the acceleration observed experimentally.
High-performance InGaN µLEDs integrated onto flexible substrates
(University of Waterloo, 2026-10-02) Gavirneni, Pranav
Flexible optoelectronic systems, from wearable biomedical devices to conformal displays, demand light sources that combine high efficiency with mechanical compliance. Indium Gallium Nitride (InGaN) microLEDs (µLEDs) are a leading candidate, yet two constraints have limited their adoption on flexible platforms: lateral scaling into the few-micrometer regime often incurs severe efficiency losses, and high-quality III-nitride epitaxy requires high-temperature growth on rigid substrates. This thesis shows that the size-dependent efficiency degradation of scaled µLEDs is largely governed by process-induced surface recombination that can be systematically suppressed. A device-centric framework is developed to decouple electrical injection, internal radiative efficiency, photon extraction, and mechanical integration, enabling the dominant loss mechanisms to be isolated and addressed in a deliberate hierarchy. Optimized ohmic contacts to n-type Gallium Nitride (n-GaN) and p-type Gallium Nitride (p-GaN) establish a low-loss electrical baseline for interpreting size effects. Sidewall-induced Shockley–Read–Hall (SRH) recombination is identified as the primary scaling-limited loss and is quantified through a practical critical diameter, critical diameter (dcrit), separating bulk- and surface-dominated regimes. Combined chemical sidewall repair and polymeric encapsulation reduce the surface recombination velocity (vs) from∼ 3×10^3 cm s^−1 to below 10 cm s^−1, with model extrapolation predicting a shift in dcrit from 43 µm to below 0.1 µm; experimentally, no size-dependent efficiency penalty was observed over the measured diameter range of 6 to 17 µm. With surface losses suppressed, geometry-driven extraction strategies become viable: perforated device architectures introduce internal sidewalls to enhance light extraction without a recombination penalty. Fully optimized devices are then transferred from sapphire onto polyethylene naphthalate substrates (PEN) via a photonic-assisted process while preserving reverse leakage below 100 fA and wall-plug efficiency (WPE) (∼ 28%). Backside potassium hydroxide (KOH) texturing of the exposed n-polar Gallium Nitride (GaN) raises the peak WPE of 5 µm devices to∼ 31%, with less than 5% variation under bending at an 11 mm radius of curvature (ROC). Overall, the results show that surface and interface physics set the practical performance limits of scaled InGaN µLEDs. The framework developed here provides guidance for engineering efficient, mechanically compliant light sources for high-resolution displays and biophotonic systems.
Decision-Focused Learning for Surgical Resource Allocation
(University of Waterloo, 2026-10-02) Li, CHUNRUI
Hospitals allocate operating-room (OR) capacity before future surgical demand is fully known, so they must rely on demand predictions when making resource-allocation decisions. A conventional predict-then-optimize approach trains a demand predictor for statistical accuracy and subsequently optimizes against its point forecast. We study whether the predictor should instead be trained directly for the quality of the OR allocation it induces. We formulate a contextual, multi-period OR-capacity allocation problem and develop a decision-focused learning approach that trains the demand predictor through the downstream allocation objective. We first establish consistency of empirical decision-risk minimization. We then analyze a two-class specialization to characterize the difference between decision-focused training and accuracy-based training using squared loss. We show that under suitable conditions, decision-focused training is decision-consistent, whereas squared-loss training targets the conditional mean and can incur a strictly positive decision regret that persists even with unlimited data. The analysis identifies capacity scarcity, reward–urgency asymmetry, and demand uncertainty as key drivers of this gap. Extensive numerical experiments confirm our analytical results. Using five years of data from a public hospital, decision-focused training reduces out-of-sample regret from 3.69% to 1.98%, eliminating 46.1% of the regret incurred by the accuracy-based predictor.