ABSTRACTS for CLASS of 2024 SENIOR PROJECTS 

as of March 22, 2024

These abstracts describe work that students PLAN to do in their projects; the abstracts for their final papers will generally be very different.

Nathanael Burns-Watson with Prof. Adam Kraus (Astronomy, UT Austin)

Determining the Host Stars of Planets in Binary Star Systems

An important characteristic of exoplanet demographics is a lack of planets in the range of 1.8-2.0 Earth radii known as the radius gap. The radius gap is thought to mark the distinction between Earth-like terrestrial planets and Neptune-like gas planets. Recent research has found that the radius gap may not be present among planets hosted in binary star systems. However, this research assumed that all of the planets were orbiting the brighter primary star. In many cases, the radius of the planet would be significantly different if it were orbiting the companion star. The aim of my project is to provide clarity on this matter by determining which star these planets are orbiting and revising their radius values in turn. This will be accomplished using data from the Kepler Space Telescope, using transit fitting and Markov Chain Monte Carlo sampling to determine the host stars and new planetary radii.

Summer Carver with Prof. Lydia Kisley and Prof. Kathleen Kash

Using Raman Microscopy to Understand the Phonon Modes of Semiconducting Ternary Nitrides

Heterovalent ternary nitride crystals have significant promise in their potential applications within the solid state lighting industry. These crystals are integrable with other binary nitrides, which are common within power electronics, and can diversify the semiconductor properties for such applications. The cation ratio within the lattice as well as the structural periodicity greatly impacts the band gap and lattice constant associated with these semiconductors. Raman microscopic analysis of crystal lattices detect phonon resonant modes to give insight into the chemical environment and structure, including the cation disorder, which directly impacts the functional properties. A Raman microscope is being constructed with an integrated fiber optic system between a BX60 Olympus confocal microscope, a 532 nm diode laser, a HoloSpec spectrometer, as well as an iDus 420 series CCD. The purpose of the microscope is to gain spatial understanding of the sample, allowing for a more detailed spectroscopic analysis of the crystal periodicity on the atomic level. Multiple crystals will be analyzed to understand their cation disorder within the structure: such samples include synthesized MgSnN2 and materials structurally related to LiGaO2. It is expected that analysis of these samples will reveal the cation ordering within the crystals, which contribute to the semiconducting properties, and give insight into the tunability of ternary nitrides. These results may facilitate designs of future crystal growth from detected crystal periodic irregularities, allowing for a better understanding of how lattice structures impact the properties of synthesized semiconductors. 

Nathan Henry with Prof. Jesse Berezovsky

Controlled Entanglement of Dipole-Coupled Nitrogen Vacancy Centers Through Manipulation of an External Magnetic Field Gradient

The Nitrogen Vacancy center, a spin-1 defect in the crystalline structure of diamond, is a promising qubit platform owing to its relatively long coherence even at room temperature. However, the primary difficulty of using these vacancies as qubits lies in the difficulty of being able to selectively entangle them. The vacancies must be near to each other in order to entangle via dipole-dipole coupling. However, since they exist in the same crystalline structure, the coupling between two centers cannot be easily stopped if they were produced too closely together. One possible solution to this problem is by applying a variable external magnetic field to the vacancies in order to split the energies of the spin states via the Zeeman effect such that the vacancies can be coupled or decoupled externally. This study intends to investigate the plausibility of this method as well as its practicability, how large the variation of the external magnetic field required to couple or decouple the vacancies is as well as the necessary speed with which the field must be changed.

Sashvat Iyer with Prof. Harsh Mathur

States of higher topological charge for systems with odd time reversal symmetry
 
The quantum Hall effect and topological insulators are remarkable new states of matter whose origin may be traced to the topology of the mapping from the Brillouin zone to the space of Bloch Hamiltonian matrices that are used to calculate the band structure of the solid. In the quantum Hall effect the Bloch matrix is complex hermitian and time reversal symmetry is broken; for topological insulators time reversal symmetry is present and odd in the sense that repeated application of the time reversal operator on a state changes its sign. In the usual case the topological states of matter involve the simplest non-trivial topological mapping but recent examples have emerged of materials in which higher order topological charge arises for the case of broken time reversal symmetry. The purpose of this project is to generalize this work to the case of odd time reversal symmetry. 
 

Robin James with Prof. Corbin Covault

The Construction and Operation of the Auger@TA Single Hexagon Array

The Telescope Array (TA), in the northern hemisphere in Delta, Utah, and the Pierre Auger Observatory (Auger), in the southern hemisphere in Malargüe, Mendoza, Argentina, are the largest ultra-high-energy cosmic ray (UHECR) observatories currently in operation. In order to increase our statistical understanding of UHECR, efforts to compare findings between the two arrays have been proposed to achieve full sky coverage. However, due to differences in the physical mechanisms, an uncertainty between the measurements produced from the surface detectors (SD) used in each observatory makes comparisons unclear. The Auger@TA project seeks to understand and cross-validate these detectors at a hardware level using an in-situ approach. This is achieved through the deployment of a single hexagon Auger SD stations array within TA to cross-calibrate the detector response of both observatories using the same cosmic ray events. To achieve this, efforts have been put into the development of the electronics and communication system for the single hexagon array in order to be deployed without the overhead infrastructure present in Auger, while still maintaining the core functionality and process to maintain reasonable certainty that the results are representative of the Auger array.
 

Joann Jones with Dr. Craig Copi and Prof. Glenn Starkman

How Improbable is our Universe? The Uncorrelated Anomalies of the Cosmic Microwave Background

The cosmic microwave background (CMB) can be described as a snapshot of the early Universe. It provides us with essential information on the origin and evolution of the Universe. Using the CMB, we are able to place constraints on the current standard model of cosmology, LCDM. Conventionally, LCDM treats the Universe as statistically homogeneous and isotropic, i.e. the same in every location and in every direction. However, there are several unexpected features of the large scale fluctuations in the CMB that are inconsistent with statistical isotropy. These are referred to as the large-scale anomalies. Each of these anomalies has a small chance of occurring (0.01 – 1%), but are often individually excused as statistical flukes. In this project, we seek to answer the question of whether or not these anomalies are correlated. In particular, we are interested in the lack of large-angle correlation, the odd-parity preference, the quadrupole-octupole alignment, and the hemispherical asymmetry anomalies. If the anomalies are indeed uncorrelated, this provides strong evidence against the existence of statistical isotropy in our Universe.

Aman Kapoor with Prof. Lydia Kisley

High Pressure Single-Molecule Investigation of Molecular Interactions with Chiral Stationary Phase Particles
 
Many molecules are synthesized as 50/50 mixtures of two mirror-image versions, termed chiral enantiomers. Separating enantiomers is essential to drug development, as each enantiomer can have varied health effects. Chiral chromatography is used to separate enantiomers but lacks a molecular-scale understanding. Previous single-molecule studies from our lab used highly inclined and laminated optical sheet (HILO) microscopy to investigate achiral analyte interactions with chiral stationary phase particles (CSPs) at pressures of 600 psi. Results showed that 2 nm analytes reach a depth of only ~20% of the CSP diameter, despite CSP pores being 50 times larger than analytes. It is unknown whether the inability of analyte molecules to penetrate CSP pores is due to low operating pressures, which are below actual column pressures of ~5,000 psi, or physical blockage by CSP polymer coatings. In this study, we build a high-pressure single-molecule setup to determine if pressure changes analyte penetration into CSPs. We pack CSPs and analyte into optically-transparent capillary tubes compatible with fluorescence microscopy and couple the capillary tube to a pressure generator that can achieve 5,000 psi. We image the packed capillary tubes with HILO microscopy and analyze single-molecule locations with super-resolution algorithms. We hypothesize that high pressures will not cause analyte molecules to penetrate the porous particles and that the lack of analyte penetration can be attributed to the polymer coating.
 

Kevin Kennelly  with Prof. Michael Hinczewski

Exploring Gradient Descent Algorithms’ Generalizability for Various Training Models

            This project will look at the gradient descent (GD) and stochastic gradient descent (SGD) training algorithms’ ability to apply to various data models. GD is a machine learning training algorithm which optimizes a set of parameters to fit a data set by approaching a target value in a landscape determined by the closeness of the parameterized function to the target, i.e., the loss landscape. SGD is a machine learning training algorithm which calculates the gradient for batches of data to generate noise in the gradient descent. By studying the eigenvalues of the hessian matrix calculated along the path of descent, the steepness or flatness of the loss landscape can be determined. The steepness directly correlates to the training algorithm’s effectiveness at finding its target in different training models where flatter loss landscapes are more generalizable than those with sharp steepness. This project aims to determine the generalizability of GD and SGD machine learning algorithms by studying factors which affect the flatness of the loss landscape such as step size and noise.

Taige Li with Prof. Kurt Hinterbichler and Prof. Harsh Mathur

New Terms in Linearized Gravity

In his lectures on gravity, Feynman demonstrated a novel way to formulate general relativity to linear order using the principles of special relativity and a few additional well-motivated assumptions. However, using the same arguments, other theories can be constructed that are not equivalent to general relativity. Our project aims to construct these models and work out their observational implications and place experimental bounds on them based on their predictions for solar system tests of general relativity and gravitational wave observations.

Grace Metz with Prof. Johanna Nagy

Designing Neutral Density Filters (NDFs) for the Cosmic Microwave Background – Stage 4 (CMB-S4) Telescope

The CMB-S4 experiment is a ground-based cosmic microwave background (CMB) experiment that will use hundreds of thousands of cryogenically cooled, superconducting detectors. In order to properly test these detectors in the lab, it is necessary to attenuate power to them equally across frequency bands using NDFs. Additionally, since NDFs have a high refractive index, they require an anti-reflective (AR) coating to prevent stray reflections within the telescope. The goal of this project is to develop a NDF and AR coating formulation for the CMB-S4 mid-frequency bands. Various NDF and AR coating formulations will be modeled to investigate their optical properties in the mid-frequency bands, fabricated, and then tested at cryogenic temperatures for failures due to differential thermal contraction of the various components.

Xavier Moskala with Prof. Benjamin Monreal

Binary Star Timing

Modified Newtonian Dynamics (MOND) and General Relativity (GR) are two of the most prominent theories for gravity, and there has been much debate over which describes the universe more accurately. We hope to quantify the experimental feasibility of testing MOND against GR by leveraging transit time variations in wide binary star systems that are expected to arise from MOND. The particular formulation of MOND that will be used is quasilinear MOND (QUMOND) as it has demonstrated better accuracy in its predictions than the original MOND formulation. In completing such work, an understanding of the lightcurves that can be measured from binary star systems and precision fitting to those curves will be instrumental. Utilizing noise approximations from telescopes such as Kepler, Tess, and JWST, we will attempt to reach a conclusion on what star systems, if any, have the ability to produce experimentally significant signals of transit time variation due to QUMOND. The continuous testing of our best theories that describe the universe is essential to learning more about the world around us, and, in this way, this work will serve an important role in describing the ability we have to test two of these theories as they relate to astrophysical phenomena.

Jonathan Willcutt with Prof. Michael Hinczewski

Counterdiabatic Control for Thermodynamic Computation

Automated image generation has been a huge success for AI, but it has also been a secret success for statistical mechanics. Underlying the techniques of stable diffusion are fundamental concepts of statistical mechanics, and thus, some of its techniques might be expected to improve it. Images must be formed slowly to ensure samples look “reasonable,” or in probabilistic terms, adiabatically moved along a sequence of equilibrium distributions. Counterdiabatic control allows for rapid changes of ground state without nonequilibrium excitations. It will be applied to a small ensemble to spontaneously form images arbitrarily quickly. The mathematics will be constrained to physically implementable operations. 

Zhi Yuan Yu with Prof. Xuan Gao

Flexible 2D Electronics

Atomically thin two-dimensional (2D) materials have many novel properties that can be exploited for device applications. One unique aspect of 2D materials is their high flexibility compared to conventional bulk crystals, opening up new opportunities in flexible electronics with high performance. The main goal for the project is to build and study flexible 2D semiconducting devices (e.g. field-effect transistor) based on atomically thin 2D semiconductors such as transitional metal dichalcogenides (e.g. MoS2), or monochalcogenides (e.g. InSe). Sub-10nm thin 2D semiconductors will be exfoliated onto flexible substrates such as polydimethylsiloxane (PDMS) and metallic contacts will be fabricated to build field-effect transistors. Their electronic behaviors will be measured and studied against the effect of stretching/bending the substrate.