ABSTRACTS for CLASS of 2021 PHYSICS SENIOR PROJECTS
as of April 19, 2021
These abstracts describe work that students PLAN to do in their projects; the abstracts for their final papers will generally be very different.
Ryan Buechele with Prof. Jesse Berezovksy
A Renormalization Group Theory Approach to Ordered Phases in Music
The organization of sounds into music has always been a fundamental part of human experience, and the desire to understand exactly how this organization arises motivates the entire field of music theory. Explanations of musical ideas among musicians are often informed by historical experience, but Berezovsky [1] has shown that musical harmony can also be described in terms of an analogy to the thermodynamics of physical phase transitions. In the same way that physical systems tend to minimize energy and maximize entropy, music can be thought of as being governed by a trade-off between dissonance and the complexity of the music. The model in [1] uses a mean-field approximation to describe the interaction between the multitude of tones in a musical system; however, this project plans to utilize renormalization group (RG) theory as a less crude way to handle the many degrees of freedom in the system. Additionally, RG theory allows for careful exploration of the phase diagram of the system to observe how the system depends on its parameters and dimensionality, as well as close investigation of the critical behavior near phase transitions. The goal of this new approach is to develop further insight into the behavior of the model and demonstrate the strength of this physical analogy by connecting the results to known systems of tuning and pitch organization.
[1] Berezovsky, J. (2019). “The structure of musical harmony as an ordered phase of sound: A statistical mechanics approach to music theory.” Science Advances, 5(5), eaav8490.Todd (Tao) Cheng with Prof. Giuseppe Strangi
Utilizing Organoids to Test for LITT (Laser Interstitial Thermal Therapy) Parameters
Laser Interstitial Thermal Therapy (LITT) is a surgical technique that is currently being developed as an alternative to current cancer treatments for several brain cancers. Glioblastoma is one of these treatable cancers, and is the current focus of this project due to the difficulty of targeting it with more conventional techniques such as radiation treatment or chemotherapy. Due to exciting new developments in organoid research there is potential for the utilization of organoids in order to simulate LITT treatment to a high enough standard that tests can be conducted to both improve the quality of future surgeries and also allow for LITT to be useful in a wider range of brain cancers. This project proposes a study on organoids to improve LITT’s effectiveness by injecting gold nanoparticles in the treated area with ultrashort laser pulses. Confocal fluorescence microscopy will allow to examine the damages on the organoid caused by LITT. It could also be utilized to serve as a preliminary test for hypothetical methods of enhancing LITT that would be more difficult in an in vivo model, and thus allow for research labs in the future to provide more initial evidence for the success of their treatment.
Benjamin Cheung with Prof. Robert Brown
Prototyping an MRI Dynamic Shim Coil
Magnetic Resonance Imaging (MRI) is a common medical imaging technique that is rapidly advancing with respect to imaging speed and quality. One issue with current technologies is that a localized difference in magnetic susceptibilities, such as between air and tissue, abruptly changes the induced field and distorts the data. An additional set of shim coils could partially compensate for this inhomogeneity, wherein the current density and discrete wire pattern required to sufficiently rehomogenize the field are reverse engineered in real time using a target-field approach. In partnership with industry, we will construct a prototype of existing shim coil design for knee imaging, so as to evaluate the efficacy of the approach and assess the extent of potential complicating factors such as unwanted coupling with surrounding coils. We will then revise and improve the shim coil design.
Skylar Danhoff with Prof. Charles Rosenblatt
Surface Patterning for Liquid Crystal Skyrmion Studies
Skyrmions are untangle-able, vortex-like knots in a vector field medium. Although originating in high energy physics, they have also been observed in various condensed matter systems. The anisotropic properties and length scale advantages of liquid crystals (LCs) provide a particularly friendly medium for probing the nature of skyrmions. We seek to create LC skyrmions in specific sizes and locations with the LC 4-n-pentyl-4- cyanobiphenyl (5CB), which has a natural tendency to conform to a patterned alignment layer and also forms a room temperature nematic phase. Our final product will be cells consisting of two glass slides, each coated with the polyamide SE1211 as an alignment layer – this material allows us to control not only the azimuthal orientation of the LC but the polar orientation as well – patterned specifically to create skyrmions and sandwiching a layer of 5CB. As such, our preliminary work primarily involves testing the integrity of our SE1211 supply and refining our SE1211 patterning technique. First, we must show that our SE1211 is homeotropic (the molecules are in a vertical orientation normal to the surface). Second, we must show that we can manipulate the tilt of the SE1211 molecules by appropriately baking and mechanically “rubbing” the SE1211 layers in a single direction. In both of these steps, we can measure the birefringence and optical retardation of light passing through completed cells to calculate the “pre-tilt” of the SE1211 molecules. Provided we can prove the SE1211 is an appropriate alignment layer for our purposes, we can then use an atomic force microscope to nano-pattern the SE1211 and scribe half skyrmion (N=1/2) lattice surface patterns. Ultimately, we hope this technique will allow us to control the size and structure of skyrmions (from microns to a few hundred nanometers) so as to eventually better understand their behavior and energetics.
Samuel Ehrenstein with Mahdi Bayat (Electrical Eng. & Computer Sci.)
Ultrasonic Microvessel Imaging Using Machine Learning
Ultrasound imaging provides an accessible, low-cost method of imaging blood microvessels in patients, which can be used for diagnosis and treatment of various conditions such as pediatric kidney disorders. At present, however, ultrasound imaging usually requires a contrast agent such as gas microbubbles to be injected into the patient. This increases the cost and risk of the procedure, making it less likely to be performed, and since most conditions are easier to treat the earlier they are diagnosed, this means patients are at greater risk of more serious complications. We propose using a machine learning approach in order to obtain comparable results to current methods such as fast iterative-shrinkage thresholding (FISTA), without the need for a contrast agent. We believe this can be done by training a neural network to reconstruct the spatiotemporal signal from the blood vessels (microvasculature) given a time series of sampled ultrasound frames (a movie). We have already succeeded in this task on simulated data, and now intend to expand our results to real (in vivo) data. Our plan is to first refine our MATLAB simulation to better simulate the physical properties of animal tissue, and then train a network on this simulated data to reconstruct blood vessels in the in vivo data.
Adam Fisher with Prof. Giuseppe Strangi
Control of Spontaneous Emission in Epsilon Near Zero Metamaterials
Cavity quantum electrodynamics is the study of photons and electrons, and their interactions within cavities of various materials at the nanoscale, in which quantum effects dominate. The results can become even more interesting when paired within metamaterials. A metamaterial is an artificial engineered nanostructure smaller than the incident light that interacts with it. They can display a multitude of fantastic properties not present in naturally occurring materials, such as near field enhancement, plasmonics, and epsilon-near-zero (ENZ) behaviors, and all that can be achieved with basic meta-surfaces. A metamaterial that drives its complex dielectric coefficient and phase of reflected light to zero for a specific bandwidth is an ENZ, which as a consequence displays many other interesting qualities. It has been demonstrated that an open-faced meta-surface with ENZ properties displays an enhancement of spontaneous emission of quantum emitters. We postulate that a cavity made of these metamaterials would display an even greater enhancement due to interaction and transfer of energy between the surfaces that make up the cavity. I intend to demonstrate this behavior by nanofabricating an ENZ-Polymethyl methacrylate (PMMA) and quantum emitter-ENZ cavity. Then measure the spontaneous emission rate of the quantum emitters using a photoluminescence spectrometer, and compare it to cavities of a simpler design and their respective open-faced designs. Meta-surfaces and cavities of metamaterials have been postulated to be the next step in advanced computing components and being able to further shrink their size. Cavity QED is on the cutting edge of optics and plasmonics research and will drive the future of nanotechnology for the next decade.
Benjamin Goldberg with Profs. Harsh Mathur and Gary Chottiner
Incorporating LIGO Detector Data Analysis into Computational Methods Curricula
First proposed by Einstein in 1916, the confirmation of the existence of gravitational waves by the LIGO collaboration opened countless new avenues in the fields of physics and astronomy and inspired a great deal of excitement throughout the whole of the scientific community. We believe that the data accumulated and the analysis employed by the collaboration, all of which has been made available on LIGO’s website, could be used in a pedagogical context, giving students of physics an opportunity to develop computational and analytical skills using the same methods and techniques as the researchers working at the cutting edge of the field. In this project, we will seek to construct a framework through which analysis of LIGO data could be incorporated into undergraduate-level computational methods curricula so as to be challenging but accessible to students. In order to accomplish this, we will develop processes using the appropriate computational software that will allow us to detect signs of black hole mergers within the LIGO data, revise these processes to a level where they would be comprehensible to a mid-level undergraduate student, and consider how they would best be used as a component of a wider educational program in computational physics.
Noah Kamm with Prof. Harsh Mathur
The Three Body Problem in Quantum Mechanics
The three body problem is a notoriously intractable problem in classical and quantum mechanics. In one dimension with delta function interactions between the particles however the problem may be exactly solved [1]. Recently it has been pointed out that the delta function is not the most general form that the contact interaction can assume in one dimension [2]. In this project we will explore whether the three body problem in one dimension remains soluble for short ranged interactions beyond the known case of the delta function. Cold atom systems provide an arena in which the physics of quantum particles in one dimension can be experimentally studied.
[1] D. Mattis, “The Many Body Problem. An encyclopedia of exactly solved models in one dimension” (World Scientific, Singapore, 1993).
[2] Thompson, Jones-Smith, Mathur and McKee, “Contact interactions and Kronig-Penney Models in Hermitian and PT symmetric Quantum Mechanics”, J. Phys. A51 , 495204 (2018).
[3] I. Bloch, J. Dalibard and W. Zwerger, “Many-body physics with ultracold gases”, Rev. Mod. Phys. 80 , 885 (2008).
Jared May with Prof. John Ruhl
Developing Millimeter-wave Scattering Surfaces for Sidelobe Control in CMB Telescopes
CMB-S4 will be a new generation of millimeter-wave telescopes utilizing a multiple-mirror design enclosed in a “cabin”. Simulations show that specularly-reflecting interior cabin walls lead to high-contrast sidelobes at angles that could hit the strongly-emitting ground during observations. CMB signals are much fainter than the millimeter-wave emissions of noise sources such as the ground so reducing the potential of these stray photons from leaking into the detector is critical. Cabin walls with some combination of scattering and absorption will greatly reduce the sidelobe contrast. This scattering and absorption is achieved through physically manipulating the material’s surface and/or applying coatings of additional materials. In addition to properly scattering and absorbing millimeter-wave photons, this material must be mechanically robust enough to endure conditions found at the South Pole. Materials will first be modeled and simulated using ray tracing software. Small samples of these materials will be fabricated and then tested using a 90GHz Gunn oscillator and diode detector mounted on a swing-arm receiver to map angular response and experimentally determine specular reflectivity, scattering, and absorptivity.
Andrew Maytin with Prof. Lydia Kisley
Expansion Microscopy Using Force
Expansion microscopy is an optical imaging method in which a specimen is expanded prior to imaging to enable super-resolution microscopy. However, the current method of using swellable hydrogels and osmotic force is limited by the amount of water the hydrogel can contain, and furthermore dynamic imaging is impossible. We aim to use tensile force to perform expansion microscopy to mechanically stretch samples. Our goal is to design a stretcher device compatible with a microscope and determine which gels have ideal mechanical properties, with the ultimate goal of producing super-resolution images using this new expansion microscopy technique.
Hannah Messenger with Prof. Lydia Kisley
Modeling the Single Molecule Microscopy Detection of Iron Corrosion
(project started in January, 2020, revised for fall 2020)
A super-resolution single molecule microscopy technique will be developed for the study of corrosion nucleation and spread at a metal interface in situ at millisecond time and nanometer length scales. Single molecule detection of corrosion is achieved using “turn on” fluorophores which become fluorescent at corrosion sites and can thus be observed through microscopy. To accompany this technique a model of the experimental system will be developed in MATLAB. Corrosion of a thin iron film will be modeled as a stochastic cellular automaton with transition rules governing corrosion pit initiation and growth. The interaction of diffusing fluorophores in the corrosive environment will be layered over the cellular automaton to simulate fluorophore turn on events as would be captured experimentally. This model will be used to elucidate experimental results and the combination of model and experiment will allow for a better understanding of corrosion in its earliest stages.
Hernan Rincon with Prof. Idit Zehavi
The Nature of Splashback Galaxies and their Connection to Galaxy Assembly Bias
Understanding the connection between galaxies and dark matter halos is essential to conceiving a framework for cosmological structure. Galaxies may be used to trace halos in the act of determining large scale dark matter distribution, provided that appropriate parameter dependencies are taken into account. The most prominent of such parameters is halo mass, but secondary contributions, including halo formation time and environment and termed together as halo assembly bias, must also be considered. The physical processes behind halo assembly bias have been studied at length, but less researched are the mechanisms behind galaxy assembly bias, the dependency of baryonic matter clustering on dark matter halo properties. One proposed explanation lies in the role of splashback galaxies, defined as galaxies traveling or having traveled through a larger host halo and which are found outside the host’s virial radius. This project will investigate what effect splashback galaxies, previously suggested to affect halo assembly bias, may have on galaxy assembly bias. Galaxy samples will be taken from the state-of-the-art Millennium Simulation with the inclusion and exclusion of splashback galaxies, and the resulting halo occupation distributions will be compared. Variance in the distributions will provide insight into the extent that splahsback galaxies affect galaxy assembly bias.
Gopal Sundaram with Prof. Cyrus Taylor
Hadley Cell Width Variability
Hadley Cells are a set of global atmospheric circulation cells that are situated above the Tropics and Equator of the Earth. Hadley cells, along with the other convection cells, are composed of hot air rising 10-15 km above the Earth’s surface, where radiative cooling cools the air to the point where the air becomes neutrally buoyant. The rising hot air pushes the now temperate air away from the Equator, with the Coriolis effect pushing the wind eastward and creating a West-East jetstream. The now cool, dry air sinks at 30° N for the Northern Hadley Cells, and 30° S for the Southern Hadley Cells. To complete the Cell, temperate winds rush South and West to the Equatorial area, creating an East-West jetstream due to the Coriolis Effect. The air rushes South due to the vacancy in the Equatorial air left by the rising hot air, and the convection cell is complete. The Hadley Cells are instrumental in creating the different climate regions on Earth’s surface, as the regions at the poleward edges of the Hadley Cells are Earth’s driest regions, due to the major influx of the aforementioned cool, dry air. Although Hadley cells are crucial in explaining the features of Earth’s climate, the dynamics of the drivers of the Hadley Cell are unclear. Along with our lack of understanding on the dynamics, experimental weather data suggests that the Hadley Cells have been expanding past their normal width, extending further than the 30° N and S that is expected. Over the past decade, climate scientists and physicists have started to investigate into the drivers behind the Hadley Cell Width Variability, and there is a growing consensus that our CO2 emissions and the radiative forcing produced by the emissions play a significant role in this problem. This project, therefore, is a computational study of the Hadley Cell Width Variation, and its potential relation to increases in radiative forcing and rising global mean temperature. We will use Carbon Emissions data, radiative forcing data, projections for global mean temperature, localized temperature and Hadley convection cell data to create a potential correlation between Hadley cell width and climate change.
Pengbo (Ricky) Wang with Prof. Charles Rosenblatt
Rayleigh-Taylor Instability at Two-layer and Three-layer Immiscible Liquid Interfaces
We investigate the Rayleigh-Taylor (RT) instability between immiscible liquids in the cases of 2-layer, 1-interface, and 3-layer, 2-interface with magnetically controlled initial conditions. A density gradient opposite of Earth’s gravitational acceleration is achieved through magnetically levitating a highly paramagnetic aqueous mixture of manganese chloride tetrahydrate, surface-tension-controlling surfactant, and fluorescent dye. The lower density fluid is fulfilled by low magnetically permeable hexadecane. The setup circumvents the issue of introducing unwanted perturbations into carefully controlled initial conditions. Analysis is carried out by studying time evolution of the arc length from video recordings of the substrate cell. We then move on to investigate interface behavior when mass is introduced into the parameter space, by means of dispersing micron-sized colloids at the interface. This illustrates the mechanisms of material transport in the context of RT instability.
Shawn Yoshida with Prof. Lydia Kisley with Will Schmid & Anuj Saini
Characterizing the Simulated Anomalous Diffusion of Proteins in Relation to the Nanoporous Structure of Extracellular Matrix-Relevant Hydrogels
Drug delivery requires therapeutics, such as proteins or RNA biologics, to diffuse through the nanoporous structure of the extracellular matrix. To enable the efficient delivery of drugs, both the nanoporous structure of this hydrogel environment and the diffusion of the drug must be understood in depth. Fluorescence microscopy data of the anomalous diffusion of fluorescently-labeled proteins BSA and fibrinogen in binarized images of a model hydrogel, at low, medium, and high concentrations will be simulated. Conventional methods are unable to quantify both nanoscale structure and diffusion, but these limitations will be overcome by correlating stochastic fluctuations from the diffusion of simulated BSA and fibrinogen in a super-resolution light microscopy technique known as “fluorescence correlation spectroscopy super-resolution optical fluctuation imaging” or “fcsSOFI.” This technique has been shown to accurately characterize Brownian diffusion, and was expanded to also characterize anomalous diffusion. The simulated datasets will be analyzed with fcsSOFI, which can quantify both the ensemble averaged and local anomaleity and diffusion coefficients, along with the size, shape, and frequency of nanopore structures. When simulating Brownian diffusion of proteins under conditions based on our experimental signal-to-background ratio, microscope parameters, and diffusion characteristics, we expect to observe subdiffusive behavior in smaller pores due to confinement. Further, Delauney triangulation can be applied to calculate pore sizes, should show agreement between the simulation and ground truth. Combining the pore size characterization with fcsSOFI’s ability to quantify local anomaleity will allow us to relate the subdiffusivity of simulated proteins with pore size. These findings can help inform drug-delivery applications where biomolecular or nanoparticle therapeutics must diffuse through the extracellular matrix.