Current Program Research

Tiaraoluwa Adesoro

Advisor: Hannah Shoenhard

Investigating the role of NCLs (Neuronal Ceroid Lipofuscinoses) genes in various cell types 

Neuronal Ceroid Lipofuscinoses(NCLs), sometimes called Batten Disease, are a family of at least 8 genetically separate childhood neurodegenerative lysosomal storage diseases from excessive accumulation of lipofuscin. This disease causes symptoms such as blindness, seizures, loss of motor ability, and premature death. However, not much is known about the precise mechanisms underlying Batten disease at the cellular level, especially in stem cells and glial cells, which support neurons. In this lab we will be investigating CLN3 and CLN7, two NCL disease genes using the fruit fly Drosophila melanogaster as a model. 

Previous data from the lab shows that, male Drosophila die more rapidly as a result of oxidative stress when CLN3 is knocked down in their intestinal stem cells. We plan to repeat this experiment and look at the effect of CLN3 knockdown on female flies when under oxidative stress. CLN3 is also expressed in cortex glia, but its function in these cells is unknown. Therefore, we also plan to investigate the effect of CLN3 knockdown in glia. CLN7, meanwhile, is expressed in perineurial glia, which form part of the blood-brain barrier and regulate sleep. To follow up on these findings, we will investigate the effect of oxidative stress upon sleep when CLN7 is knocked down in the perineurial glia. These results are important in contributing to the understanding of the mechanism of Batten disease, and contributing to research in neurodegenerative diseases. 

Srijanee Basu

Advisor: Monica Chander

Co-culture Analysis of SoxR Activation and Actinorhodin Production in Streptomyces coelicolor

Streptomyces coelicolor is a soil-dwelling filamentous bacterium known for producing a wide range of bioactive secondary metabolites, including the benzochromanequinone metabolite actinorhodin (Act). Among its notable chemical properties is its redox-active nature. SoxR is a redox-sensing transcription factor that is highly conserved across a range of bacteria, both enteric bacteria (like Escherichia coli, where SoxR protects the bacterium from global oxidative stress) and non-enteric bacteria (like S. coelicolor, where SoxR appears to serve a different physiological role). 

Previous research conducted both within the lab and outside the lab point towards SoxR functioning to maintain redox homeostasis in S. coelicolor by sensing and responding to Act levels. However, it is important to recognize that S. coelicolor shares its soil habitat with a variety of antibiotic-producing organisms, and that its SoxR can be activated by several exogenous redox-active compounds, including phenazines produced by Pseudomonas aeruginosa. Thus, studying S. coelicolor in monoculture may overlook key regulatory and metabolic dynamics that are only triggered in the presence of external redox stressors or competing species.

Thus, this summer, the Chander lab aims to conduct a series of co-culture experiments to investigate interspecies interactions between various mutant strains of P. aeruginosa and S. coelicolor. These experiments will focus on SoxR activation, Act production, and the broader implications of these processes for how S. coelicolor copes with environmental stress.

Caitlyn Caya

Advisor: Bárbara Bitarello

Investigation of Adaptive Evolution In TAS2R14 and TAS2R38 Bitter Taste Receptor Genes in Primates 

In primates, bitter taste perception is thought to function as an adaptive response to avoid ingesting toxic compounds. This response is mediated by type 2 taste receptors (T2Rs) encoded by a family of G protein-coupled receptors known as TAS2Rs. In mammals, the TAS2R family is dynamic, with variations in gene number observed across species.  In humans, the TAS2R gene family has 25 functional TAS2Rs distributed across 3 chromosomes (chromosome 5, 7, 12). Among these, T2R14 can recognize over 150 bitter agonists, while T2R38 binds to 23 known ligands and influences individuals' perceptions of phenylthiocarbamide (PTC). Variation in PTC in taste perceptions can be linked to polymorphisms in TAS2R38. Recent studies have shown that both T2R14 and -38 are expressed extra-orally, including in airway smooth muscle, the gastrointestinal tract, and innate immune responses, suggesting there is a broader physiological role beyond taste perception. Gene clusters of genes are present in chromosome 7 and 12 containing closely related TAS2R genes. Hence, we will use TAS2R38 to be the representative gene from chromosome 7 and TAS2R14 for chromosome 12, as they are both extensively researched genes.

Prior work in the Bitarello Lab identified site-specific signatures of adaptive evolution occurring pervasively in these genes in Primates, yet many questions remain. Does adaptive evolution (AE) target specific functional regions, such as ligand-binding pockets? There is a lack of agreement in the literature regarding what amino acid residues compose the ligand-binding pockets. To address this, we will conduct a thorough review of available information about binding sites in TAS2Rs. Another open question is whether accelerated evolution in these genes is (a) primate-specific, (b) present beyond primates (e.g., other mammals), and (c) restricted to certain lineages more than others within primates and mammals. 

This summer we aim to analyze TAS2R14 and -38 across primates and non-primate mammals. Orthologs will be identified using TOGA2, followed by estimation of the nonsynonymous to synonymous distribution rates ratio (dN/dS) using PAML (site models) and HyPhy. Ultimately, we hope to shed light on the ligand binding mechanisms in these receptors and trade-offs involved in having such a promiscuous receptor (TAS2R14). TAS2R38, a much more fine-tuned receptor, will serve as an appropriate contrast in our attempt to understand these trade-offs.

Catherine Chang

Advisor: Thomas Mozdzer

Quantifying Heritability of Phragmites australis Stomatal Characteristics Under Varying Levels of CO₂ and N 

Phragmites australis, also known as the common reed, is a globally distributed invasive wetland grass and a proposed model species for studying plant responses to environmental changes. Past studies have found that elevated carbon dioxide (eCO₂) and nitrogen enrichment (Nenr) influence carbon fixation through changing both morphological and physiological traits, such as stomatal size, stomatal density, and photosynthetic rates. However, it was not determined whether these changes to eCO₂ and Nenr were heritable (G), plastic (E), or due to heritable plasticity (GxE). 

My research will investigate and differentiate the contributions of genotype and phenotypic plasticity in response to global change. To examine stomatal trait heritability, P. australis pedigreed families of genotypes were exposed to two levels of CO₂ and two levels of N conditions at the Smithsonian Global Research Wetland (Edgewater, MD) in 2025. Epidermal peels will be taken from pressed leaf samples to reveal stomata impressions, which will then be measured and counted using a microscope at 50x magnification. Given the high number of experimental replicates (n=384), my project also aims to develop high throughput tools that will automate the quantification of stomatal characteristics. This will be done by evaluating different AI programs to most efficiently and accurately measure stomata length, width, and density. Overall, the first outcome of this project will be the development of a pipeline to characterize stomatal characteristics, followed by AI tools to help automate the process. My findings will also provide fundamental insights on both the heritability and plasticity of these traits to near future global change, which has implications for carbon fixation and ecosystem resilience of coastal wetlands.

Zilu (Lucy) Deng

Advisor: Tamara Davis

Analysis of DNA Methylation Profiles and Gene Expression in Dnmt1 Mutant Mice

Cells regulate gene expression through DNA methylation by adding a methyl group to cytosine at CpG sites (cytosine followed by guanine). Methylation typically silences gene expression and is carried out by DNA methyltransferases (Dnmts). Among the three different DNA methyltransferases, Dnmt1 is responsible for maintaining methylation patterns after DNA replication, adding methyl groups to the newly synthesized strand by copying the patterns from the template strand. Because proper concentration of gene products are important for development and survival, abnormal methylation levels can alter gene expression levels and lead to diseases and even death. The T. Davis lab analyzes mice that contain a mutation in the intrinsically disordered domain of Dnmt1, resulting in reduced ability to maintain DNA methylation post-replication, which leads to global hypomethylation. Since the consequences of the Dnmt1 mutant have not yet been fully studied, we are interested in looking into how the loss in methylation affects gene expression.

Previous studies in T. Davis’s lab have shown that homozygous mutant Dnmt1 mice exhibit perinatal death. However, the embryos show no morphological differences between wild-type and mutant. Therefore, we hypothesized that the perinatal death may be due to neurological defects or general viability issues. To test this, I selected three differentially expressed genes that are involved in either the central nervous system or in neuron differentiation pathways (all upregulated in the Dnmt1 mutant) and aim to assess their methylation levels across different developmental stages and tissues. Given the reduced ability of Dnmt1 to maintain methylation in the mutant, I hypothesize that methylation levels will decrease at these upregulated genes. To assess methylation profiles, I will use bisulfite mutagenesis, which converts unmethylated cytosines to uracils while leaving methylated cytosines unchanged. I will then use RT-qPCR to validate the differential expression of these three genes and see if they exhibit any temporal or tissue specific variations.

Nethmi Dharmasena & Raina Teter

Advisor: Adam Williamson

Research was performed in the lab of Adam Williamson, Department of Biology

Azury Gomez-Ulloa 

Advisor: Alison Weber

Wing Mechanosensory Responses to Flight Perturbations in Manduca sexta

Proprioception, the ability to sense one's body position and movement, is necessary for all organisms to effectively interact with their environments. However, how these signals are transformed into actions remains poorly understood in most systems. Insects offer an opportunity to investigate this through their small nervous systems, predictable motor pattern, and accessible sensory structures . Sensory feedback is especially critical for stability in flight. Campaniform sensilla (CS) are microscopic mechanosensory structures integrated in the wings that detect wing bending during flight and contribute to flight stability. We examine CS in the hawkmoth Manduca sexta because of its large body size and relatively slow wingbeat frequency, which make it easier to track behavior and record neural activity than in smaller insects. We record neural activity using extracellular electrophysiology, where a small probe is inserted into the wing nerve to record action potentials. Wingbeats are simulated using a motor that drives periodic wing motion, and air gusts are delivered to mimic destabilizing conditions. Investigating responses to these gusts will contribute to our understanding of how sensory systems detect and respond to unexpected challenges during flight. This work could help uncover general rules about how the nervous system processes proprioceptive information and uses it to guide behavior.

Asal Mogharehdehkordy

Advisor: Monica Chander

Investigating the Role of Actinorhodin and SoxR in Redox Homeostasis in Streptomyces coelicolor

Streptomyces coelicolor is a soil-dwelling bacterium that produces a wide variety of secondary metabolites. One of these metabolites, actinorhodin (Act), is a redox-active, blue-pigmented antibiotic that is primarily produced during the stationary phase, when cells reach high population density. Although the physiological function of Act remains incompletely understood, it has been proposed to play a role in cellular redox balance under stress conditions. One model suggests that during the stationary phase, nutrient limitation slows oxidative phosphorylation, leading to electron leakage from the electron transport chain and the production of reactive oxygen species (ROS). Under these conditions, Act may function as an alternative electron sink, helping to reduce oxidative stress. The redox-sensitive transcription factor SoxR is activated during Act production and may contribute to this process by regulating genes involved in maintaining Act redox cycling.

For my project, I will investigate how Act and SoxR contribute to maintaining redox homeostasis in S. coelicolor. I hypothesize that Act is produced in response to energetic and oxidative stress and SoxR facilitates this protective process. To test this hypothesis, I will compare Act-producing and Act-deficient strains, as well as strains lacking SoxR, under conditions that disrupt cellular respiration. Using compounds that interfere with oxidative phosphorylation, I will examine changes in growth, stress sensitivity, and Act production. This research will provide insight into the physiological role of Act and the function of SoxR in redox regulation. Ultimately, these findings may improve our understanding of how antibiotic producing bacteria adapt to metabolic stress and maintain cellular homeostasis.

Aurora Gutterman-Johns

Advisor: Thomas Mozdzer

Elemental Analysis on Root Samples of Phragmites australis from Genotypes Exposed to Different Global Change Conditions. 

Phragmites australis, the common reed, is known to influence wetland ecosystem stability, and is becoming increasingly important as an invasive model organism to study the effects of global change. Coastal wetlands are known to play a critical role in the many ecosystem components and the stability of coastal regions which are increasingly threatened due to global change factors. P. australis has the capacity to rapidly evolve to near future global change, which makes it an ideal species to study. The aim of this work is to understand trait heritability of plant traits that influence carbon cycling in wetland ecosystems. I am analyzing root tissues from individual genotypes from an exposure experiment to ambient CO₂ + ambient nitrogen (N), ambient CO₂ + elevated N, elevated CO₂ + ambient N, and elevated CO₂ + elevated N. Roots samples from genotypes exposed to all 4 conditions (n=384 total) will be uniformly ground to be prepared for elemental analysis of C and N content on an elemental analyzer. Samples will also be combusted to determine ash content allowing me to calculate the energetic construction costs of plant genotypes under different global change conditions. 

Hannah Jensen-Sizelove

Advisor: Thomas Mozdzer

Coastal wetlands act as a buffer between land and sea, providing essential ecosystem services such as storm protection, carbon sequestration, and pollutant filtration. Currently, these ecosystems are threatened by anthropogenic-driven global change. Phragmites australis, or the common reed, is a widespread invasive grass that is increasing its range in North American coastal wetlands. Given its cosmopolitan distribution, Phragmites has become a model organism for invasive species and global change research, At a long-term field site at the Smithsonian Global Change Research Wetland, Phragmites has been shown to undergo rapid evolution under exposure to elevated nitrogen enrichment and elevated CO₂ concentration. Our lab sampled Phragmites from this population, and from another population in Maryland, to create a quantitative genetic common garden at Bryn Mawr College. Over the past three years, our lab has created 51 pedigreed families using the NC1 crossing design resulting in 234 parental genotypes and 362 F1 genotypes allowing us to estimate the heritability of plant traits across two populations. We plan to measure traits on all 596 genotypes in order to estimate quantitative genetic parameters like narrow sense heritability. Identifying which traits are heritable allows us to better understand the effects of rapid evolution in an uncertain world. 

This summer, I will work to determine the heritability of three distinct traits, both aboveground and belowground, including specific leaf area (SLA), leaf chlorophyll content (LCC), and root shear strength. SLA, the ratio of the leaf surface area to its dry mass, serves as a strong predictor for plant growth. LCC informs a plant’s photosynthetic capacity and potential for carbon fixation. Root shear strength, is a proxy for belowground biomass allocation, and provides insight into a plant’s ability to prevent erosion and stabilize shorelines. These traits contribute to Phragmites’ functional role, and understanding their degree of heritability helps determine how evolution will impact their ecology. 

Nirali Lakhani

Advisor: Alison Weber

Quantifying behaviors of hawkmoths to understand compensatory movements when exposed to destabilizing perturbations

During flight, many insects rely on visual, mechanosensory, and other sensory inputs to respond to environmental changes. While mechanosensory input is known to compensate for slower processing of visual inputs during flight, little is known about the role of wings in providing this mechanosensory feedback. To fill in this gap, we are examining behavioral responses to sensory inputs to the wings, using hawkmoths as a model organism. Hawkmoths offer key advantages because of their relatively large size and accessible nervous system, allowing us to understand the role of wings in executing precise compensatory movement in response to destabilizing perturbations.

We will record high-speed videos of tethered hawkmoths under normal flight conditions and under destabilizing conditions with a wind gust. Then, we will analyze the movements of keypoints using markerless 3D tracking software. This experiment will help us improve our understanding of the mechanosensory role of the wings in response to destabilizing perturbations. 

Dakarai Lindsay

Advisor: Thomas Mozdzer & Brian Donnelly

Decomposition rates of Phragmites australis litter

Globally, salt marshes are recognized among the world’s most productive blue carbon ecosystems, exhibiting higher rates of carbon uptake and long-term carbon storage than terrestrial ecosystems. Salt marshes serve as a model ecosystem for studying global change and eco-evolutionary responses, helping to understand carbon influx to soils. Ecosystem carbon emissions are primarily driven by the balance between primary production and respiration, much of which is driven by the decomposition of organic matter. As an ecosystem engineer, Phragmites australis is a model organism for investigating global change factors such as elevated CO₂ and nitrogen enrichment, as well as rapid eco-evolutionary processes. Changes in heritable trait variation of these plants as a result of global change may alter carbon cycle processes, positively or negatively, thereby altering salt marshes' carbon storage capacity.

This project aims to analyze the decomposition rates from an exposure experiment that was run in 2025 at the Smithsonian Global Change Research Wetland.  Our lab used the Tea Bag Index to understand the influences of plants on decomposition rates of organic matter under current and near-future global change scenarios. We understand that global change can induce rapid evolution, resulting in genotypic differences within Phragmites australis populations. It is unknown if global change acts as a selective pressure on plant traits that influence decomposition rates, which have the potential to increase or decrease carbon storage. Our study utilizes both green tea and red tea as standardized substrates to assess these decomposition dynamics. Bridging eco-evolutionary insights with measures of decomposition processes, this work seeks to clarify whether these notable carbon sinks will continue to remain a major carbon sink or become a carbon source in the future. 

Emma Loller

Advisor: Thomas Mozdzer

Examining heritable and plastic trait variation in elemental content of of Phragmites australis leaves under global change

Coastal wetlands face increasing pressure from global change factors such as rising CO₂ and nutrient enrichment. The common reed, Phragmites australis, is a model organism for studying plant invasion and physiological responses to global change in coastal wetlands. Research has shown that Phragmites undergoes rapid evolution in response to these conditions, with shifts in genetic diversity and genotype-level trait variation. Elemental content of carbon (C) and nitrogen (N) have previously been shown to demonstrate trait heritability. This summer, I will analyze 384 individual plants that were exposed to elevated CO₂ and N in a field study at the Smithsonian Global Change Research Wetland in Edgewater, MD, in 2025. Elemental analysis will be performed on Phragmites australis leaves to evaluate both the heritability and plasticity of these traits, and results will be connected to trait variation to evaluate whether C and N content are related to heritable trait expression in this species.

Bianca Perez Ouhirra

Advisor: Hannah Shoenhard

The effects of autophagy knockdown in the a'β' medial mushroom body in relation to synaptic plasticity in Drosophila melanogaster

In the brain, as neurons fire together the connections between them become stronger through an increase in the size of the synapses between each cell, a process known as synaptic plasticity. Synaptic plasticity is theorized to be responsible for learning and memory in humans, with synapses returning to their baseline state during sleep. During sleep deprivation and aging, synaptic plasticity becomes dysregulated, over-strengthening synapses. The mushroom body is the learning and memory center of the Drosophila central brain, with the a'β' m region contributing to memory consolidation during wake periods. Last summer, we found that brain-wide synaptic stability is supported by autophagy in a'β' m neurons. We observed changes in the synaptic marker bruchpilot (Brp) as a reference for synaptic size. We found that with an autophagy knockdown of 3 days synapses across the entire central brain potentiated significantly. Longer knockdown of autophagy in these neurons also caused a decrease in the brain's natural autoflorescence, which may indicate a loss of the autofluorescent metabolite NADH, a key factor in cellular energy production.

This summer session, I intend to continue to research the relationship between a'β' m neuron activity, autophagy, and the brainwide synaptic potentiation phenotype. I will repeat the complete autophagy knockdown I performed last summer with additional controls and alternative tools to confirm my previous results. Additionally, we were curious as to what subtype of autophagy may be responsible for the synaptic phenotype. There are two main pathways by which autophagy may be performed: one leading to lysosomal degradation and recycling of autophagosome contents and the other leading to the extracellular secretion of the waste. To analyze this I will perform knockdowns of exclusively lysosomal and secretory autophagy genes and compare the results. The goal of this experiment is to understand the significance of each type of autophagy in the potentiation phenotype exhibited.  Finally, I aim to explore the autofluorescence phenotype by directly measuring NADH. 

Rosa Ramirez Almanzar

Advisor: Gregory Davis

Research was performed in the lab of Gregory Davis, Department of Biology

Sambridhi Shrestha

Advisor: Tamara Davis

Paradoxically Downregulated Genes and their Role in the Perinatal Death of Dnmt1 Mutant Mice 

Mammals are complex organisms. Our complexity is the result of the regulated expression of over 20,000 genes, which is tightly maintained by various forms of gene regulation - one of which is DNA methylation. DNA methylation is an epigenetic mechanism which involves the addition of a methyl group at the C5 position of a cytosine in CpG dinucleotides to form 5-methylcytosines. DNA methylation silences gene expression either by recruiting repressor proteins or inhibiting the binding of transcription factors to DNA. The enzymes necessary for establishment and maintenance of these methylation patterns are DNA methyltransferases 3a and 3b (Dnmt 3a & 3b), and Dnmt1 respectively. Improper DNA methylation leads to various cancers, autoimmune diseases as well as neurological and developmental disorders. The T.Davis lab studies a mouse strain with a mutation in the intrinsically disordered domain of Dnmt1 to investigate how decreased Dnmt1 activity in mice alters DNA methylation patterns and gene expression.

Homozygous mutant Dnmt1 mice have reduced Dnmt1 activity which results in a genome-wide hypomethylation. This results in perinatal death with no morphological difference in comparison to wildtype mice. It is hence expected that genes silenced by methylation will be upregulated in Dnmt1 mutant mice; however, preliminary data from the lab showed paradoxical downregulation of a subset of genes in the mutant brain samples. Over the summer, I will be studying two of these downregulated genes; Scn5a and Sptb. Scn5a is a gene that encodes a sodium channel responsible for initiating heartbeat by mediating the depolarizing phase of action potential in myocardial cells. Sptb is a gene that encodes a structural protein which is a primary constituent of erythrocyte plasma membrane cytoskeleton, essential for hematopoiesis. Dysfunction of either could lead to the perinatal death of Dnmt1 mice without significant morphological difference. Additionally, I will also be studying one of the upregulated genes of the Scn family: Scn4a which is responsible for mediating the depolarizing phase of action potentials and muscle contraction. My goal for the summer is to identify the mechanism behind this paradoxical downregulation and to characterize these genes’ methylation patterns using Bisulfite Mutagenesis which converts unmethylated cytosines to uracils. Additionally, I will perform RT-qPCR using RNA from various developmental stages to assess temporal changes in the differential expression of these genes.

Maeve Slemko

Advisor: Hannah Shoenhard

Energetic Regulation of Sleep and Wake Memory Consolidation in Drosophila melanogaster 

Sleep is one of our most important biological functions. It replenishes energy stores and clears toxic waste buildup. Sleep also enhances memory consolidation, the process of moving a memory into long-term storage. Memory consolidation is highly energy intensive. In the Shoenhard lab, we hypothesize that sleep allows the brain to operate in an "energy saving" mode, making it more energy-efficient to consolidate memory during sleep. In the model organism Drosophila melanogaster, the common fruit fly, memory consolidation can occur during either wake or sleep. Both forms of memory are controlled by neural circuits in a brain region known as the mushroom body. We wish to determine how these specific memory circuits utilize energy stores, such as lipids, to show how energetic strategies for memory consolidation differ during sleep vs wake. We will use transgenic Drosophila lines to test how selectively hyperactivating or inhibiting α’β’ a/p neurons and α’β’ medial neurons influence lipid metabolism. Additionally, we will test whether natural activation of these circuits via sleep and wake memory training assays influences lipid metabolism. We measure lipid metabolism by starving the flies and measuring their time to death and quantify lipid levels through thin-layer chromatography (TLC). I hypothesize that stimulating α’β’ a/p neurons will increase triglyceride levels, while inhibiting these neurons will decrease them, and stimulating α’β’ medial neurons will decrease triglyceride levels, while inhibiting them will increase them. These findings will contribute to understanding the differences between wake and sleep memory, with implications for learning, diet, and sleep.

Monica Anderson

Advisor: Eugenia Vasileiadou

Synthesis Optimization and Ag Doping of 370/390 CdSe Magic Size Clusters

The small size of cadmium selenide nanocrystals—typically less than 100 nm—allows for unique optical and electronic properties distinct from bulk materials that can be rationally tuned during synthesis, as recognized by their Chemistry Nobel prize in 2023 and rendering quantum dots commercial success in products like QLED televisions. Despite this, much remains uncharted about their smaller < 2 nm variants termed CdSe magic size clusters. In contrast to typical nanocrystals, MSCs exist at specific thermodynamically stable structures as intermediates, while larger nanocrystals have a distribution of sizes. My research over the summer will investigate the synthesis, growth mechanism and characterization of a specific family of CdSe MSC with absorption peaks at 370 nm and 390 nm. First, I will define which synthetic parameters lead to the synthesis of CdSe MSC versus larger quantum dots by tuning several reaction parameters. Next, I will characterize the optical properties of all my nanocrystal products using absorption and photoluminescence emission spectroscopy to study the growth mechanism. I will measure the exciton lifetime to determine whether the source of photoluminescence derives from surface defects, trap states or—most ideally for imaging applications—the CdSe core. I will conduct Ag doping of my CdSe MSC to study the evolution of the optical properties, which is hypothesized to improve the emission of the MSC which is typically less pronounced than that of larger nanocrystals. Finally, through cyclic voltammetry, I will investigate the band structure of the MSC, gaining further insight into the eƯects of doping on the electronic properties of the nanocrystal intermediates. Mapping out the growth and properties of CdSe MSC will provide synthetic and structural guidelines in improving the synthesis of nanocrystals, overall promoting further their imaging applications.

Anabelle Bahadur

Advisor: Jonas Goldsmith

A Sustainable Approach for Producing Hydrogen Fuel Through the Synthesis and Polymerization of Transition Metal Complexes

The current thermochemical processes used to produce hydrogen gas utilize fossil fuels which, in turn, releases carbon dioxide into the atmosphere causing detrimental harm to the planet. The Goldsmith Lab aims to produce hydrogen gas using renewable energy sources. In order to do this, water photoreduction takes place with the help of catalytic bimetallic polymer complexes. These systems act as photosensitizers (PS) and electron relay (ER) molecules, where the ER quenches the excited PS system through an electron transfer mechanism. The proximity of the PS and ER systems are imperative to the rate of electron transfer, thus by the synthesis and sequential layering, or polymerization, of ruthenium, rhodium, and iridium complexes containing vinyl bipyridine ligands the electron transfer rate increases. Additionally, these ligands yield both electrochemical and photoactive properties which allow for the effectiveness of these complexes to be studied using cyclic voltammetry (CV) and chronoamperometry (CA) as light energy enters the system causing electron transfer to take place from the PS to the ER molecule. Ultimately, the aforementioned research aims to offer a sustainable approach for producing hydrogen fuel by the electropolymerization of transition metal complexes. 

Lucie Burgess

Advisor: Eugenia Vasileiadou

Electrochemical and Structural Investigation of Optimized Electrode Components for Fluoride-Ion Batteries

The Chemistry Nobel prize in 2019 was awarded for the “development of lithium-ion batteries” that are essential in our daily lives. In recent years, there has been an increase in developing new battery technologies due to changes in global supplies of lithium, along with rising financial and environmental costs. Fluoride-ion batteries (FIBs) are more environmentally friendly and cost-effective alternatives to lithium-ion batteries as fluorine is much more abundant than lithium. Fluorine’s high electronegativity makes it a great element in redox reactions, which is how energy is stored in batteries. My project will focus on investigating how the anode and cathode electrode components in these batteries individually and collectively impact the battery performance.

 First, I will evaluate potential candidates as anode materials for FIBs. To do this, I will test a series of binary metal fluorides MFx in a symmetrical cell configuration: M (metal) / liquid electrolyte / MFx using electrochemical methods such as cyclic voltammetry and galvanostatic charge cycling. Once the most promising anode is determined, I will move forward on investigating a potential cathode material for FIBs. The FeSb₂O₄ schafarzikite crystal structure is a cathode compound of interest for FIBs. I will work on the solid-state synthesis of FeSb₂O₄, where I will use powder X-ray diffraction for structural characterization and to confirm synthesis of the targeted product. Finally, I will conduct electrochemical studies with FeSb₂O₄ cathodes in a coin cell battery. This would be accomplished by creating FeSb₂O₄ slurry electrodes and fabricating coin cells to measure their electrochemical properties to understand how effective the FeSb₂O₄ cathode and fluorine ion are at redox reactions, therefore their potential as battery materials. Overall, my research will provide structural and electrochemical evidence on how to improve the anode and cathode of FIBs.

Rukmina Dhakal

Advisor: Bill Malachowski

Synthesis of Nitrogen-Containing Cyclic Structures with Quaternary Centers 

Heterocyclic rings and quaternary carbons are very common in drug molecules. Nitrogen-containing rings provide polarity and hydrogen bonding, increasing solubility in aqueous solutions and increasing intermolecular force interactions with molecular targets, thereby increasing binding affinity. Quaternary carbons promote selective interaction, reducing side effects and off-target binding. This summer, I will work on synthesizing heterocyclic rings and tetrahedral carbons using low-cost, abundant starting materials to improve the efficiency of creating small-molecule drug candidates. The research uses a desymmetrizing alkene hydroamination reaction where an amine adds across an unsaturated C-C bond to form a C-N bond to convert benzoic acid derivatives into saturated indole and quinoline rings (heterocyclic rings) with bridgehead quaternary carbons, catalyzed by copper and lithium metals. Copper, lithium, and benzoic acid derivatives are low-cost, abundant starting materials, while copper poses less harm to the environment compared to other alternatives, such as palladium and/or nickel. This research could make a profound impact in the field of medicinal chemistry by providing valuable information about efficiency in drug making. 

Leena Dzemaili

Advisor: Eugenia Vasileiadou

Investigating Novel Synthetic Routes for Silver Telluride Quantum Dots

Quantum dots are so small that their optical properties can be tuned simply by changing one property: size. Aleksey Ekimov, Louis Brus, and Moungi Bawendi were awarded the 2023 Nobel Prize for the discovery and synthesis of these unique materials whose brightness and tunable properties make them valuable for a wide variety of display applications. Despite these exceptional qualities, however, many quantum dots are synthesized with heavy metals such as cadmium or lead, hindering their broader applications. 

Silver chalcogenide (S, Se, Te) quantum dots are non-toxic and heavy-metal-free. Their unique ability to absorb and emit light in the infrared region renders them promising candidates for bioimaging applications. However, as these materials are less studied than heavy-metal-based quantum dots, the question remains how to best optimize their synthesis to ensure strong optical properties in the infrared spectral region. 

To explore this question, I will investigate a wide parametric space of synthetic conditions that produce high quality, monodisperse silver telluride quantum dots.  I will examine the nature of both silver and telluride precursors. Using a Schlenk line, the influence of an inert atmosphere will also be examined. In parallel, I will investigate how the formation of intermediate silver nanoparticles affects the quality of the final Ag₂Te product. The mechanism of Ag₂Te quantum dot growth, as well as optical properties, will be monitored through absorbance and photoluminescence emission spectroscopy. Photoluminescence properties will also be studied using a fluorimeter with a photomultiplier tube detector, permitting access to the infrared region. Finally, to assess the brightness of my synthesized Ag₂Te quantum dots, I will measure the photoluminescence quantum yield. In all, my research will define new synthetic routes for silver and silver telluride nanocrystals, uncovering the routes with optimal infrared properties.

Áine Himsl-Fenz

Advisor: Patrick Melvin

Investigating the Capabilities of a Benzothiazolium SCF₃ Reagent in Various Oxime Rearrangement Reactions

Fluorine is a particularly valuable element due to its unique ability to alter important pharmaceutical properties including a molecule’s stability, lipophilicity, and ligand affinity. Because of this, developing methods of the inclusion of fluorine and fluorine-containing functional groups has been of great interest in the field of organic synthesis. To this end, a benzothiazolium salt has been developed to incorporate an interesting thiotriofluoromethyl (SCF₃) group into various organic molecules. The reagent, known as BT-SCF₃, is bench-stable and requires only mild conditions for synthesis of various SCF₃-containing compounds. This research uses the synthesized BT-SCF₃ reagent to explore the addition of medically significant trifluoromethylthio (SCF₃) groups to different chemical compounds via a rearrangement mechanism. Through the combination of the BT-SCF₃ reagent with amidoxime substrates, we aim to synthesize a novel class of SCF₃-containing formamidines.  

Madeline Le

Advisor: Ashlee Plummer-Medeiros

Microscopic Analysis of Lipid Binding with LetB in Escherichia Coli

Gram-negative bacteria, such as E. coli, are surrounded by a double membrane barrier and lipid trafficking to the outermost membrane is important for protection, specifically antibiotic resistance. The LetAB complex facilitates movement of lipids from the inner to outer membrane. This project focuses on the periplasm-spanning tunnel protein LetB and understanding how it binds to membranes and mediates lipid trafficking.  LetB acts as a tunnel to transport lipids to the E. coli outer membrane likely utilizing its seven pore-lining loops (PLLs) which line the interior hydrophobic tunnel. However, little is known on how or if these PLLs affect membrane binding. This project involves deleting each of the PLLs to observe how membrane binding of the mutated protein differs from the wild type version through a novel microscopy-based assay. The proteins will be expressed in E. coli, purified with affinity chromatography, and characterized through size-exclusion chromatography and SDS-PAGE. For membrane binding experiments, proteins will be added to a supported lipid bilayer, followed by the addition of free flowing vesicles containing fluorescently tagged lipid derivatives. The membrane binding affinity of each protein will be determined by quantifying vesicle-binding in the presence of varying concentrations of protein – our hypothesis is that the wild type protein will have the strongest interaction with the membranes and that the LetB mutation with deletion in PLL 1 will have the weakest interaction with membranes because that is where the first lipid interaction occurs in LetB. This research aims to understand how each PLL affects membrane binding and is important in understanding the overall lipid trafficking mechanism of the LetAB system. If this mechanism can be better understood, novel antibiotics may be designed to more effectively inhibit the transport of lipids between bacterial membranes.

Emma Li

Advisor: Bill Malachowski

Enantioselective desymmetrizing hydroamination to synthesize complex all-carbon quaternary stereocenters in nitrogen-containing heterocyclic rings

A large part of research in drug development focuses on searching for new synthetic tools and more efficient ways to create complex molecular architectures. Heterocyclic rings and tetrahedral carbons are two structures that have been commonly found in recent, successful drugs. Heterocyclic rings, especially those with nitrogen atoms, are polar and capable of hydrogen bonding, which is useful for solubility in aqueous solutions, and ultimately promoting drug-protein binding affinity. Tetrahedral carbons are capable of creating selective interactions, which minimize unwanted side reactions with non-target proteins. While bioactive molecules that contain both structures in the form of fused heterocyclic rings with bridgehead quaternary chiral carbons exist, there still lacks an efficient method that can enantioselectively construct these molecules. 

The current proposed method synthesizes symmetrical cyclohexadiene structures from Birch reduction-alkylation reactions of inexpensive benzoic acid derivatives and then desymmetrizes them through enantioselective alkene hydroamination reactions.

Similar to past research in the Malachowski lab, which has involved using nickel and palladium in the Mizoroki–Heck reaction, my work will investigate the use of chiral copper catalysts in facilitating the hydroamination reaction. Copper is an attractive alternative due to its low cost and positive environmental attributes. Using copper catalysts, I will be developing an enantioselective method to saturated benzoxazinone and 2-quinazolinone rings with quaternary carbon stereocenters. I will then work to adapt the method to the synthesis of saturated indole and quinoline rings with quaternary carbons, and more broadly, synthetic pathways to a range of nitrogen-containing heterocyclic rings with quaternary carbon substituents.

Afia Serwaa Marfo

Advisor: Bill Malachowski

Synthesis of Enantioselective Fused Heterocyclic Compounds with Quaternary Carbons through Birch Reduction- alkylation Reactions.

Heterocyclic rings and tetrahedral carbons are common components of successful small molecule drugs, as they facilitate drug solubility in aqueous solutions and create selective interactions with protein molecules, reducing unwanted side reactions, respectively. All-carbon quaternary compounds are the most difficult types of tetrahedral carbons to synthesize because there are limited efficient synthetic tools for their construction. However, fused heterocyclic compounds with bridgehead quaternary carbons are seen in large number of bioactive molecules. This highlights that the development of a synthetic tool for these compounds promises a potential acceleration in the success of small molecule drugs. The goal for my research is to explore a proposed efficient synthetic tool to enantioselectively construct  fused heterocyclic compounds with bridgehead quaternary carbons from inexpensive benzoic acid derivatives. The proposed method utilizes  Birch reduction-alkylation reactions and will develop enantioselective alkene hydroamination reactions. Birch reduction-alkylation reactions will be used to synthesize symmetrical cyclohexadiene structures, followed by desymmetrization through enantioselective alkene hydroamination reactions. The desymmetrizing step will be facilitated by a chiral copper catalyst. Computational chemistry will also  be used to model the amino-cupration step, which is suspected to be the rate and enantioselectivity determining step.   The use of a copper catalyst makes this method efficient as copper is low in cost and less hazardous. A catalyzed reaction also ensures atom economy or an efficient use of resources.  If successful, this work has a potential of expanding  methods for the synthesis of complex molecules, as well as developing more effective small molecule drugs. 

Ahlaam Musa

Advisor: Jonas Goldsmith

Synthesis of Bimetallic Transition Metal Complexes For Hydrogen Based Energy

The transition towards a low-carbon future renders current methods of hydrogen energy production unsustainable and calls for the development of environmentally friendly methods for energy production. This project seeks to produce green hydrogen energy without the formation of carbon dioxide as a byproduct. This is done through the synthesis of a bimetallic molecule that will work as a catalyst for the reduction of water using light energy, producing hydrogen gas. Previously, hydrogen production via water photoreduction has been performed using catalytic systems composed of metal complexes which act as photosensitizers (PS) and a separate electron relay (ER) molecule. The ER quenches the excited PS through electron transfer, creating a reactive species which will reduce protons to hydrogen gas. To establish an effective connection between PS and ER within a limited timeframe, the PS and ER are brought into close proximity through amide coupling, accomplished by synthesizing transition metal macromolecules, with carboxylic acid one carbon away on either PS or ER and amine on the other. This project will focus on the synthesis of complexes with ruthenium centers. This research has the potential to drive sustainable energy production, significantly reducing greenhouse gas emissions and contributing to global goals for climate remediation. The mechanism for the attachment of groups used in this project can also have further applications in different synthesis methods.

Aden Noor

Advisor: Eugenia Vasileiadou

Exploratory Synthesis and Optoelectronic Properties of Hybrid Palladium Halide Perovskites 

Hybrid halide perovskites have emerged as promising materials for next-generation solar cell applications. These hybrid organic-inorganic semiconductors are made from solution methods, rendering them a low-cost alternative to traditional semiconductors like silicon, which require vacuum methods for production in solar cells. Lead-based perovskites, with the general formula of APbX₃ for 3D structures and BA₂MA_n-1Pb_nX_3n for 2D structures (where A = methylammonium or formamidinium; X = Cl^-, Br^-, I^-; BA = butylammonium and n = 1, 2, 3, 4, … ) are the main perovskites investigated for solar cells. Replacing Pb would mitigate toxicity concerns of lead perovskite solar cells. Our research will explore the synthesis and characterization of alternative palladium halide perovskite compositions and determine if they are viable substitutes. 

Firstly, I will synthesize a series of hybrid palladium halide compounds through various solution methods (ambient-pressure synthesis, hydrothermal synthesis) and collect powder X-ray diffraction data to determine that the intended perovskite crystal structure has formed. Next, for novel compounds identified through powder X-ray diffraction, I will move forward with single-crystal X-ray diffraction measurements to derive the precise atomic structure of these novel compounds. 

Finally, I will characterize the optical properties of my synthesized samples. Currently, there is little to no data published on the absorbance and photoluminescence behavior of hybrid palladium perovskites, making this an intriguing area of research. I will use absorbance and photoluminescence spectroscopy to measure the optical properties of my compounds. Our research will create a foundation of hybrid palladium halide compounds that have the potential to be applied in affordable and sustainable solar energy devices.

Shoshana Peck

Advisor: Bill Malachowski

Enantioselective synthesis of heterocyclic compounds

Common components in many successful small molecule drugs are heterocyclic rings and tetrahedral carbons. Heterocyclic rings increase the solubility of drugs in aqueous solutions and their ability to bind with protein targets, while tetrahedral carbons make these protein interactions selective to minimize off-target interactions that could cause unwanted side effects as well as potentially increasing dissolution.

There is no general synthetic tool to enantioselectively create fused heterocyclic rings with quaternary carbons at the bridgehead, despite their prevalence in bioactive molecules, especially in nature. This research seeks to develop such a synthetic tool that uses inexpensive benzoic acid derivatives as the starting material. These derivatives undergo Birch reduction-alkylation reactions to generate symmetrical cyclohexadiene structures that are then desymmetrized through enantioselective alkene hydroamination reactions.

Chiral copper catalysts are being explored for use in the desymmetrization reaction, building on previous use of palladium and nickel chiral catalysts for enantioselective desymmetrizing Mizoroki-Heck reactions and extending other researchers’ copper-catalyzed hydroamination work to use in enantioselective desymmetrizing reactions. The copper catalysts have the benefits of being sustainable, low toxicity, and inexpensive.

It is estimated that many unique chemical structures are difficult or impossible to synthesize with current synthetic tools. Research such as this hopes to address this issue, by developing synthetic tools to create chemical structures with a wide range of applications.

Brianna Reino

Advisor: Bill Malachowski

Synthesis of Nitrogen-Containing Polycyclic Heterocycles Through Copper-Catalyzed Desymmerization Hydroamination

Heterocyclic rings and tetrahedral carbons have polarity, hydrogen-bonding capabilities, and three-dimensional shape that make them significant components of successful small molecule drugs. These intermolecular force interactions with protein targets promote drug-protein binding and increased selectivity in pharmaceutical applications. Nitrogen containing heterocyclic rings are successful in many pharmaceutical drugs because they have the correct intermolecular forces necessary for drug use. Finding new methods to access nitrogen heterocycles and tetrahedral carbons centers would facilitate the efficient and selective synthesis of these molecules. The Birch reduction-alkylation reaction creates a symmetrical cyclohexadiene molecule that can then act as the reactant in controlled enantioselective desymmetrizing reactions. This results in bicyclic nitrogen heterocycles with quaternary stereocenters. Heterocyclic rings containing quaternary carbons are found in many bioactive molecules; however, there is no general synthetic strategy to selectively control the stereochemistry in these molecules. The prevalence of quaternary carbon centers in bioactive molecules, combined with their synthetic difficulty, makes them intriguing targets for research. To address this challenge, this proposal will investigate the use of chiral copper catalysts to facilitate an enantioselective desymmetrizing hydroamination reaction in order to construct poly-heterocyclic rings with all-carbon quaternary chiral centers. Copper is an attractive catalyst for the development of this proposal due to its low cost, low toxicity, and environmental sustainability.  

Stella Scarozza

Advisor: Ashlee Plummer-Medeiros

Investigation of Charged Transmembrane Residues in the YebS Transport System

Gram-negative bacteria, such as E. coli, are surrounded by a double membrane system, which both protects the cell from external threats and facilitates cell survival. The inner membrane is composed of a phospholipid bilayer, while the outer membrane is asymmetric, with an outer leaflet of lipopolysaccharides and an inner membrane of phospholipids. These membranes are separated by an aqueous periplasm. The phospholipids that compose these membranes are synthesized within the inner membrane, which necessitates lipid transport across the periplasm – an energetically unfavorable process due to the hydrophobic nature of the lipids. This process is facilitated by several protein systems within the double membrane, one of which includes the inner membrane protein YebS (also known as LetA).
This project will focus on better understanding the role of charged transmembrane residues in YebS lipid-trafficking function. YebS contains several highly conserved charged amino acid residues within the transmembrane region. In order to discern their importance, several residues will be replaced with the neutral amino acid alanine. Mutated YebS will be purified from E. coli using affinity and size exclusion chromatography. Purified proteins will then be reconstituted into synthetic donor vesicles, which mimic the bacterial inner membrane, and an in vitro transfer assay can be performed to measure fluorescently tagged lipid movement as facilitated by the mutated YebS. Based on previous computational work, it is hypothesized that the mutation of the Glutamate 161 and Arginine 362 will be most detrimental to lipid transfer. This research will further our understanding of the lipid transport system in E. coli, and therefore open doors to the design of novel antibiotics to specifically inhibit bacterial growth.

Ramona Shekhar

Advisor: Bill Malachowski

Synthesis of heterocyclic rings with tetrahedral carbons with benzoic acid

Synthesizing heterocyclic rings with quaternary chiral carbon centers is essential for developing small-molecule drug therapies because they facilitate drug-protein binding affinity and minimize unwanted side effects due to their selective interactions with protein targets, in addition to promoting small-molecule dissolution. For my research, I will be conducting a synthetic pathway via Birch-alkylation reduction reactions of benzoic acid derivatives with various alkylating agents such as iodocyclohexane and 1-bromo-3-chloro-propane to create cyclohexadiene structures. We hope to expand on the enantioselective desymmetrizing Mizoroki-Heck reactions and copper-catalyzed hydroaminations conducted by chemists in the past.

Madelyn Soden

Advisor: Patrick Melvin

Investigating the Potential of Sulfone Iminium Fluoride Reagents in the Synthesis of Carbodiimides

The Melvin group has previously studied sulfone iminium fluorides (SIFs), a sulfur reagent, that adds fluorine onto organic molecules. The SIF reagent has proven itself to be useful with fluorination chemistry, and the goal of this summer is to explore how SIFs can be used in other ways. One idea of interest is the synthesis of carbodiimides using the rearrangement of secondary amidoximes. Certain carbodiimides have been found to be useful in tissue engineering. One example is collagen scaffolding, where the carbodiimide acts as a crosslinker. The SIF shows potential as a synthetic reagent as it is favored to facilitate the rearrangement chemistry necessary to from the desired carbodiimide product. To accomplish the formation of this valuable functional group, acid chloride oximes will be synthesized starting from various aldehydes. A primary amine will be reacted with the acid chloride oxime, forming an amidoxime. When the SIF reagent is combined with the amidoxime, the hydroxyl group of the oxime is converted to a viable leaving group which facilitates the rearrangement. This methodology will allow for a wide variety of carbodiimides to be synthesized and isolated. 

Sodtugs Tsengelbayar

Advisor: Bill Malachowski

Enantioselective Synthesis of Nitrogen-Containing Heterocycles via Birch Reduction-Alkylation and Aziridination

Many successful drugs contain nitrogen heterocycles and quaternary carbon stereocenters, but synthesizing these structures with enantioselective control is a challenge. This summer in the Malachowski lab, I will work toward this goal starting with a Birch reduction-alkylation, followed by LAH reduction and carbamate or sulfamate formation to set up an aziridination. The aziridine intermediate will then be functionalized to access a range of nitrogen-containing structures, including imidazolines and related heterocycles. The core focus is developing the method for fused heterocycles with quaternary centers, with the broader goal of establishing a general route to complex drug-like structures from simple starting materials.

Zoe Woon

Advisor: Ashlee Plummer-Medeiros

Gram-negative bacteria, such as E. coli, protect themselves against antibiotics or other threats and pathogens through the maintenance of their asymmetric double membrane. Protein systems such as LetAB are important for membrane integrity and the transport of lipids across the aqueous periplasm. This system is also a part of the mammalian cell entry protein family, which has been linked to virulence factors in other species, such as M. tuberculosis. Studying the structure and function of these protein systems has the potential to contribute to combatting antibiotic resistance, of which more than 2.8 million cases are diagnosed in the United States each year, by providing new therapeutic targets. Specifically, this research aims to examine the interactions between transmembrane protein LetA and periplasm-spanning LetB, primarily within the bacterial inner membrane. These interactions are studied through computational simulations at both an atomistic and coarse-grained level. The dynamics of these equilibrated systems are simulated through the application of classical physics force fields in the programs NAMD and GROMACS, for all-atom and coarse-grain simulations respectively. Lipid-protein interactions, protein-protein interactions, and lipid movement of these simulations are analyzed using ProLint2 and Visual Molecular Dynamics programs. The analysis of wild-type LetA and LetAB simulations may potentially highlight frequent or long-lasting interactions, which can be further studied with the mutation of the relevant wild-type residues to alanines and comparisons of these variants to the wild-type simulations. Overall, this research will provide a detailed perspective of the LetA and LetB interface and contribute to our understanding of membrane regulation, which is important for bacterial virulence and the design of novel drugs to combat antibacterial resistance.

Fatma Ayad

Advisor: Elizabeth Dinella

Liveness: Measuring the Refreshability of LLM Benchmarks

As large language models are trained on an increasing amount of web corpora, there is also a growing risk that widely-used benchmarks are included in this training data, compromising their competency in providing unbiased evaluations. This phenomenon, also known as benchmark data contamination, leads to rapid benchmark saturation, where the models appear to perform well due to data memorization rather than true capability. Many recent works have proposed the concept of live or refreshable benchmarks that depend on continuous evolution and updating to remain challenging to LLMs regardless of data exposure. However, the property of benchmark “liveness” remains a novel notion that has yet to be formally defined and properly quantified. 

Our work focuses on investigating liveness as a benchmark property and exploring the methods to quantify an LLM benchmark’s capacity to refresh and remain resistant to memorization. Firstly, we plan to examine whether techniques from membership inference research can be adapted to estimate the extent to which benchmark examples have been memorized by a model and use these signals to develop a metric for benchmark liveness. Secondly, we will examine recent benchmarking efforts, including work on live and unsaturated evaluation frameworks, and evaluate how effectively the proposed metric captures differences in benchmark longevity and susceptibility to contamination. 

By introducing a measurable notion of liveness, we aim to improve the design of our benchmark RubberDuckBench to better combat data contamination. We also plan to evaluate state-of-the-art benchmarks such as SWEBench and DeepSWE by using this metric to assess their resistance to memorization and benchmark saturation. In conclusion, our work aims to contribute a framework for assessing benchmarks’ ability to still provide meaningful evaluations as large language models continue to rapidly evolve.

Huda Imran

Advisor: Elizabeth Dinella

Improving and Expanding RubberDuckBench: A Benchmark for Coding Assistants 

Software Engineers are increasingly relying on chat-based LLM tools for coding tasks, yet it is not clear how well these models truly understand code. RubberDuckBench is a benchmark consisting of code reasoning questions and detailed evaluation rubrics, for analysing the performance of AI coding assistants. 

This project expands RubberDuckBench in three key ways. First, we broaden the benchmark by adding samples and introducing more robust rubric criteria beyond omissions and hallucinations, incorporating factors such as relevancy and succinctness. Additionally, we conduct a higher-level analysis of LLM performance across programming languages. Currently models perform best on Java and worst on C++, and we aim to determine whether this reflects model capability or question difficulty. Finally, we include build scripts and a framework to compile and run each project, this enables us to conduct execution-based evaluation alongside text-based analysis.

Together these improvements will allow RubberDuckBench to more deeply and precisely evaluate AI’s understanding of code, contributing to the goal of improving LLM tools for software engineering.

Emily Lu

Advisor: Elizabeth Dinella

Agentic AI for Code Reasoning

Despite the widespread adoption of AI in software development, state-of-the-art large language models (LLMs) still struggle to answer natural language questions about complex codebases. While standalone LLMs underperform, we have also found that equipping an advanced model like Claude Opus 4 with basic Unix tools often increases computational costs without improving correctness. To overcome these limitations, we plan to explore a more sophisticated framework that pairs LLMs with robust program analysis and dynamic testing tools. Instead of relying on static environments, we plan to implement sample-specific build scripts that allow the system to autonomously generate, execute, and evaluate test cases. Furthermore, we intend to integrate delta-debugging via tools like C-Reduce to automatically isolate minimal bug triggers from these generated test suites. We expect this multi-layered approach to significantly improve accuracy, as providing the agent with execution-driven feedback and minimized test cases allows it to validate its reasoning and isolate root causes dynamically. Finally, to ensure strict validation of our framework's outputs, we plan to deploy three independent rubric applicators for each LLM output. Through these methods, we aim to bridge the gap between static model generation and dynamic program execution.

Eman Rasool

Advisor: Elizabeth Dinella

Automated Application of Rubrics for Evaluating LLMs and Humans 

The rapid expansion of Large Language Models (LLMs) has created an urgent need for scalable and reliable evaluation frameworks. Rubrics are the gold standard for providing structured, consistent grading in both professional benchmarking and educational settings, such as AP Computer Science Free-Response Questions (FRQs). However, manually applying these rubrics is labor-intensive and difficult to scale, while current Automatic Short-Answer Scoring (ASAS) techniques often fail to capture this nuance by relying on surface-level semantic similarity to reference answers.  This research aims to bridge that gap by developing a technique for the automated application of detailed rubrics. We plan to explore the adaptation of concept-based scoring and neural approaches to directly interpret rubric criteria, moving beyond simple lexical overlap. To evaluate the effectiveness of this technique, we will test it against the outputs of various LLMs within the RubberDuckBench framework, a multilingual benchmark of real-world contextualized questions derived from GitHub pull request comments, and human-graded AP Computer Science FRQs . Our goal is to achieve a higher degree of agreement with human graders than current state-of-the-art methods. By creating a more interpretable and automated evaluation system, this project contributes to the development of trustworthy AI and more efficient educational assessment tools.

Lorraine Nguyen

Advisor: Michael Wehar

Advancing the AlgoArt Creator Studio: Enhancing User Experience for the Creation of Drawing Algorithms

In the field of creative computation and digital art, accessible software tools are crucial for enabling users to generate artwork through code. AlgoArt is an established platform developed to facilitate this intersection of computer science and visual design. Within this ecosystem, the Creator Studio serves as the primary environment where users directly interact with algorithms, parameters, and generated images to produce algorithmic artwork.   

While the Creator Studio has a strong technical foundation built by multiple contributors over time, its current workflow presents usability challenges. The processes of editing code, manipulating parameters, and generating images are somewhat fragmented. This disjointed user experience interrupts the creative process and creates a barrier for users attempting to experiment with their designs.   

For summer research, we propose to unify and optimize the user experience within the Creator Studio by redesigning its core interfaces. The initial phase of this project will focus on evaluating and improving the code editor and parameter interface. By leveraging JavaScript and modern web interface design principles, we will streamline how users transition between writing algorithmic logic and adjusting visual settings, ensuring a more cohesive environment.   

The subsequent phase of this project will explore and implement new software features for artwork storage to enable iterative design processes. Specifically, this involves creating features for saving and comparing generated outputs during active sessions. By developing a save log of recent artworks and enabling side-by-side reviews, users will be able to track their design history, which will connect the Creator Studio’s experimental workspace to the platform’s gallery and review system. 

Ultimately, we not only aim to enhance the technical usability of the Creator Studio but also seek to improve AlgoArt as an educational tool. By establishing a more organized and unified user workflow, our summer research will improve user experience and accessibility of the platform for students and creators, enabling better experimentation, education, and creation in the field of algorithmic art.

Chang Sun

Advisor: Elizabeth Dinella

SolidityAFL: A High-Performance Coverage-Guided Fuzzer for Ethereum Smart Contracts

Smart contracts power billions of dollars in decentralized applications, yet vulnerabilities in their code continue to cause significant financial losses. Fuzzing, a technique that generates random inputs to stress-test programs, has proven effective at finding bugs in traditional software, and has been increasingly applied to smart contracts. However, recent studies reveal that existing smart contract fuzzers suffer from low throughput and limited coverage feedback, limiting their ability to find bugs.

We present SolidityAFL, a smart contract fuzzer built on LibAFL, a modular fuzzing framework written in Rust. Unlike prior fuzzers that rely on older C++ frameworks or custom execution engines, SolidityAFL leverages revm, a standalone Ethereum Virtual Machine also written in Rust, to achieve high transaction throughput. Our employs type-aware input generation that respects Solidity's Application Binary Interface, and can seed its corpus with real-world on-chain transaction data to begin exploration from realistic program states. We also implement a constraint system for specifying inter-function dependencies and test oracles that detect invariant violations indicative of unauthorized fund withdrawals.

We plan to evaluate SolidityAFL against existing contract fuzzers using established vulnerability benchmarks such as SolidiFI, measuring throughput, code coverage, and vulnerability detection rates. In parallel, we will extend the fuzzer's capabilities in two directions: restoring fork mode to support fuzzing directly against live on-chain contract state, and introducing multi-core parallelization to dramatically improve fuzzing throughput. We aim to build a practical security tool that helps developers identify and fix vulnerabilities before their contracts are deployed.

Emma Blersch

Advisor: Pedro Marenco

Organic and inorganic carbon abundance and isotope signatures in the lower Triassic (Spathian) Virgin Limestone of Lost Cabin Springs, Nevada 

The end Permian mass extinction was one of the most deadly and disruptive events in earth history, fundamentally changing the structure of biotic and abiotic systems. Analysis of rocks formed in the early Triassic, specifically those of Spathian age from 247 to 245 million years ago, gives insight into biotic recovery and the state of the planet after an extinction event. The Virgin Limestone member of the Moenkopi Formation in Lost Cabin Springs, Nevada has a Spathian section containing interbedded carbonates and stromatolite mounds, representing the transition period of ocean life in the early Triassic. The end-Permian saw an elevated amount of volcanic activity and release of greenhouse gases from the Siberian Traps, leading to extreme global warming and the disruption of ocean circulation, conditions reflected in the rock record, which can be deduced with carbon isotope analyses. 

As part of ongoing research into the Lost Cabin Springs mounds, my research will focus on organic and inorganic carbon abundance and isotope signatures of the collected samples. Samples were collected from a section of over 200 meters, and this analysis will compare the isotopic ratios and carbon abundances across the entire formation. To analyze the various carbon signatures of the Lost Cabin Springs samples, I will be seperating the organic and inorganic carbon from the powdered rock and using cavity ring down spectroscopy to detect isotope ratios and carbon abundance. My work with isotope profiling will provide a higher resolution of this section of the Moenkopi Formation, along with finer details concerning dating and ocean-wide conditions at the time of formation.

Jayden Coffey

Advisor: Selby Hearth

The Boreal forest is a large, global collection of ecosystems making up earth's largest land biome. The biome makes up for roughly 30% of the forested area on earth’s surface, making a significant contribution to the biodiversity of plant and animal life and supporting life. Throughout the sprawling landscapes within the biome there are significant coal deposits- some of which have served or currently are active coal mines.

Coal presents both geologic and political interests. As a “rock” it embodies a complex and diverse geochemical and tectonic history, and allows geologists to better understand the unique climate conditions of the past that allowed for its formation at least 2 million years ago and way before then. As a resource, it represents a source of fuel that has allowed for some civilizations to make advancements in modern technology and development, and thus profit off of it.

While the world continues to develop new approaches to managing extraction and energy consumption, coal continues to pose a major threat to its surroundings in its depositional environment and after extraction. Coal serves as a source of fuel when exposed to oxygen, which under the right conditions may ignite itself (combust) and progress to a coal seam fire. Coal seams fires can degrade the surrounding soil as well as spread to the surface, leading to deep remote forest fires that can go undetected.

For this reason, the large coal deposits within earth's largest forested biome must be further researched in order to develop better fire monitoring and prevention tactics. My research project will focus on exploring the geo-chemical and spatial relationship between coal and the boreal forest through literature review and GIS Mapping analysis. The goal of this project is to determine the geologic methods capable of locating coal formations within the focus area and defining the best practices for monitoring.

Julia Hancock

Advisor: Don Barber

Investigating the influence of local tidal and hydroclimate variability on groundwater levels and saltwater intrusion in Davis, NC

In low-lying coastal areas, sea-level rise due to climate change can cause shoaling and salinization of the water table in shallow aquifers. As sea-level rise has accelerated,  saltwater intrusion, overland flooding, and waterlogged ground have begun to adversely affect coastal communities with increasing frequency. These issues can be exacerbated by climate-change driven intensification of the hydrologic cycle, such as more severe droughts and heavier precipitation during storms. In many rural coastal communities, the undesirable impacts, such as roadway flooding and salt-killed vegetation, associated with these phenomena disproportionately affect lower income households with limited means for remedies and adaptation.

In order to quantify the occurrence and severity of groundwater shoaling and salinization, monitoring wells that continuously log conductivity, temperature, and water height have been deployed in the unincorporated community of Davis, in eastern North Carolina.  The current monitoring array in Davis, consisting of three shallow wells and a local tide gauge, will be expanded with the installation of three more groundwater monitoring wells during fieldwork early this summer. Data from the enhanced monitoring array will be integrated with a high-resolution digital elevation model to create a numerical model of groundwater flow in the study area. Analysis of the results will improve our understanding of how saltwater intrusion and elevated water levels impact rural coastal settlements and will inform efforts to provide adaptive solutions to increase the resilience of vulnerable communities. 

Charlize Hoeun

Advisor: Don Barber

Effects of Salinity, Inundation, and Adjacent Land-Use on Below-ground Salt Marsh Carbon Stocks in Eastern North Carolina

Saltwater marshes and wetlands serve as some of Earth’s most prominent carbon sinks, absorbing substantial amounts of CO₂ from the atmosphere faster and for longer periods of time compared to other carbon sinks. Salt marshes additionally serve as buffers that protect upland environments from storm surges. These important ecosystem services may be lost if marshes are unable to keep up with accelerated sea level rise due to climate change. Upward building of marshes through sediment deposition and thick intertwining plant root frameworks has allowed vertical accretion that has kept pace with relative sea-level rise for thousands of years, but recent anthropogenic climate change and other human impacts appear to put marsh survival at risk. These effects, including increased inundation, higher temperature and salinities, and decreased suspended sediment availability create complex and interacting challenges for marsh survival.

      One example is increased evaporation on interior marsh platforms due to high summer temperatures, which our prior research has shown leads to higher salinities, approaching levels toxic even to salt-tolerant plants. The resultant marsh die-back exposes more bare mud to heating by summer sunlight, thereby further increasing water evaporation and marsh salinification. Another anthropogenic impact reported by Himmelstein and others (2026) is the creation of marsh-edge levees that hinder interior marsh sedimentation, therefore weakening marsh resilience to sea-level rise. The proximity of marshes to cleared land areas, whether it be for farmland or residential development expansion, exacerbates the production of such levees, subsequently depriving sediment load to the same interior marsh areas experiencing plant die-back from high temperatures and salinities. This further hinders marsh build up, leading to further elevation lag behind rising sea level.

      My study will evaluate the interaction of these perturbations by analyzing below-ground carbon content of sediment from salt marsh sites with different proximities to cleared farmland, and with different saltwater inundation times. The resulting data on carbon content in soil cores provides a test of whether marsh carbon sequestration is decreasing in efficiency compared with the recent geological past (last several hundred to 1,000 years). By quantifying how eastern North Carolina marshes are responding to external global change perturbations, this study will inform the management of marshes and adaptation strategies of coastal areas moving forward.

Olivia Johnson

Advisor: Katherine Marenco

Microbialites in the Early Triassic: A lamination-scale geochemical analysis of microbial mounds

Following the End Permian mass extinction, sponge-microbialite reefs have been interpreted to be markers for biotic recovery from the mass extinction. This is thought to be due to the ecological niche which opened up for reef-building organisms, and the unique geochemical conditions of the Early Triassic oceans. Microbialites form from the buildup of laminations produced by the activity of microbial communities. Geochemical analysis of the individual layers in a stromatolite from a microbialite mound may provide insight into the environmental conditions the stromatolite lived in. Samples collected from the Lost Cabin Spring locality of the Virgin Limestone of the Moenkopi Formation in Nevada, USA will be assessed at the lamination-scale for elemental and inorganic carbon isotopic compositions. In addition, these samples will be compared to microbialite samples from the Phanerozoic. Data collected across these samples will be interpreted on the lamination-scale within individual stromatolites to understand what changes may have occurred for that sample. The data will also be compared across all sites and samples to identify any geochemical trends. Comparison across these stromatolites offers further insight into how localized conditions may impact microbialite growth, and if there are any consistent patterns across geologic time which inform the conditions under which microbialites were most successful.

Katharine (Kit) Kringel

Advisor: Katherine Marenco & Pedro Marenco

Investigating geochemical and environmental contexts for microbialite reef formation in Early Triassic marine rocks of the Lost Cabin Springs Formation

The prevailing hypothesis of microbialite reef formation in the Phanerozoic is that such structures can only form in inhospitable marine conditions, where a lack of metazoan life allows microbial mounds to dominate reef ecosystems in the absence of predation and competition. This often links microbialite formation to post-extinction environments and extreme geochemical conditions, but recent studies have suggested otherwise. In particular, robust microbialite structures formed in the Early-Triassic may provide evidence that microbialite reefs are able to form in highly productive, hospitable marine environments, and that there are different, yet-unknown mechanisms controlling their success.

This project will utilize inorganic carbon isotopes, organic carbon isotopes, and other elemental analysis to determine the geochemical conditions preserved in samples from two microbialite-bearing sections of the Moenkopi Formation in Las Vegas, Nevada. The results of the project will provide proxies for local environmental conditions, creating a geochemical profile of the section that can then be compared to other known microbialite sections to establish spatial and temporal patterns, and test common understandings of microbialite reef formation. 

Lily Muehlenhard

Advisor: Arlo Weil

Insight into ‘true’ grain-scale fabrics that produce bulk Anisotropy of Magnetic Susceptibility measurements in weakly deformed red sandstones 

Measuring the Anisotropy of magnetic susceptibility (AMS) gives critical insight into how the internal structure of a rock deforms as a response to an externally applied tectonics stress. Notably, while the AMS of a sample’s bulk composition can be analyzed, the AMS of individual mineral populations within the sample are unknown.  

Utilizing the Quantum Diamond Microscope (QDM), I will investigate a possible new laboratory technique that measures AMS at the grain scale.   

With the QDM, I will produce a dimensional map of AMS at the microscopic scale over a given sample’s surface area. The micro-scale result will then be compared with the macro measurement made for standard AMS core measurements. Maps of magnetic coercivity and magnetic remanence have previously been done on rock samples, but a map of magnetic susceptibility and its spatial anisotropy has not. If successful, this technique will provide a new method of internal rock fabric analysis and may illuminate grain-scale processes that give rise to a sample’s bulk anisotropy fabric. In addition to the QDM and AMS measurements, various rock magnetic experiments will be done on all measured samples to characterize the magnetic nature of the specimens being measured.

Anya Pulichino

Advisor: Pedro Marenco

It is agreed widely that the End Permian Mass Extinction was triggered by the extensive eruption of the Siberian Traps (~252 million years ago). The mechanisms of the mass extinction are still debated, with deep ocean anoxia being one of the main hypotheses. Certain elements, such as uranium, are redox sensitive, which allows us to investigate the presence of oxygen in past oceans. 

The Virgin Limestone Member of the Moenkopi Formation exposed at Lost Cabin Springs, NV, formed in the Early Triassic a few million years after the End Permian Mass Extinction. I will be using samples collected throughout the whole Lost Cabin Springs section to measure the abundances of uranium and other redox sensitive elements to investigate ocean oxygenation in the Early Triassic. To do this, I will be using Inductively-Coupled Plasama Mass Spectrometry to measure the abundances of uranium and molybdenum within each sample. Results from this study will provide insight into environmental conditions associated with possible mechanisms for the End Permian Mass Extinction. 

Ava Purser

Advisor: Arlo Weil

Finalizing an Investigation on Laramide Deformation in the North American Cordillera 

The Rocky Mountains of Western North America exist as the result of a major mountain building event that began roughly 150 million years ago. Deformation initiated with normal subduction of the Farallon plate beneath the North American plate in the Jurassic and growth of the thin-skinned Sevier fold-thrust belt that extended from Arctic Canada to Northern Mexico. Farallon plate subduction then transitioned to a flat slab geometry in the Late Cretaceous, resulting in a change to a thick-skinned deformation style across the Rocky Mountains. The macro-scale structures produced during these two mountain building phases serve as a record for the mechanics by which the ancient Farallon Plate was subducted.

The data for this research has been collected from hundreds of sites across the region where anisotropy of magnetic susceptibility (AMS) is used to measure Laramide shortening directions. Measurements of structural deformation such as vein sets and minor faults were also collected to contextualize the AMS data. In addition, paleomagnetic analysis is used to correct the data for any measured vertical axis rotations that may have occurred as a result of complex tectonic movement. The quantification of shortening directions across the region inform the kinematics of the Rocky Mountain tectonic system and can be extrapolated to interpret its broader geodynamic history. 

This summer, we will travel between the Laramide Colorado Plateau and the Southern Sevier belt of Utah to collect additional samples. My goal is to analyze and collect samples to fill current gaps in research that will ultimately constrain existing Rocky Mountain deformation models and assist in establishing a robust geologic model for the evolution of the Cordilleran-style mountain system as it relates to ancient plate boundary dynamics.

Malin Thomas

Advisor: Don Barber

Heavy metal abundance in Mercenaria clam shells and soft tissue from North Carolina and New Jersey

Aquatic bivalves such as Mercenaria provide an account of heavy metal pollution in their environments and can even display otherwise undetectable metal contamination in water. Measuring the levels of these contaminants in bivalves is also important because shellfish are a marine food resource and elevated metal concentrations in clam tissue pose health risks. Beyond pollution, normalized metal concentrations in shells have been used as proxies to reconstruct past environmental conditions such as temperature, salinity, and nutrient levels. 

Our research will investigate the degree to which patterns of metal abundance in aragonite clam shells correspond to observed metal concentrations in soft tissue from the same organism. For example, do the shells of clams with high soft tissue lead concentrations contain elevated Pb/Ca ratios? 

We are analyzing Mercenaria samples collected from coastal New Jersey and North Carolina that display varying soft-tissue metal abundances. Specifically, we will measure the concentrations of arsenic, cadmium, copper, lead, nickel and zinc in both shell and tissue using inductively coupled plasma mass spectrometry. 

In addition to comparing intra-organism tissue and shell metal concentrations, we will evaluate how metal concentrations vary among our New Jersey and North Carolina collection sites. We will compare the Pb/Ca ratios in our North Carolina shells from 2022-2026 to those from a previously published paper that analyzed shells collected between 1949-2002 from the same sites. Our study will shed light on how metal concentrations in clams vary in response to environmental conditions, and the degree to which Mercenaria shells record estuarine metal pollution. 

Taliah Thomas

Advisor: Don Barber

Microscopic and geochemical analyses of modern Mercenaria clam shells as bioindicators of recent extreme climate events in North Carolina.

Sclerochronology is the study of the recurring physical and chemical features recorded in the mineralized hard parts of organisms, as well as the rate and timing of their formations. Sclerochronological analysis of clamshells can inform research on environmental conditions and ecosystem perturbations to the clam’s estuarine habitat. As clams grow and mature, they precipitate calcium carbonate in an organic protein rich matrix, forming distinct shell increments separated by growth lines. By counting these growth lines and measuring the spacing between them, we can identify growth variability patterns and provide temporal context for environmental variability. Within the same clam shell, geochemical and isotopic analyses can reveal additional environmental information including varying temperature, salinity, and pollutant levels during the clam’s life. 

For my research I will make polished sections of live-collected shells of Mercenaria mercenaria from the central coast of North Carolina, specifically from Core and Back Sounds. The polished shell sections will be etched and stained using Mutvei’s solution, which is a mixture of acetic acid, glutaraldehyde, and alcian blue, thereby enhancing the visibility of growth lines under microscopic analysis. I am going to compare my growth increment analyses to instrument records of seawater temperature, tide level, and salinity from monitoring sites near where the clams were collected. Microsamples from paired sections of the growth-counted shells can be analyzed for stable carbon isotopic composition as well as for metal concentrations in the Park Science Geochemistry Lab. Results of these geochemical analyses will be compared with the growth increment findings, as well as other environmental information relating to hydroclimate and water chemistry in the study area.

Christine Dabuco

Advisor: Leslie Cheng

Mathematical Foundations and Applications of Arbitrage

An arbitrage opportunity refers to the ability to generate a profit without investing any initial capital. This summer, my research project will focus on exploring applications of arbitrage within a financial context and also through examining and attempting to exploit market inefficiencies in parimutuel horse races. In the context of parimutuel horse racing, outcomes are uncertain and pricing is driven by a collective behavior, which creates a unique environment in which inefficiencies may arise. 

Through this work, I will develop a rigorous mathematical understanding of arbitrage by introducing a mathematical definition and engaging with proofs of well-known arbitrage theorems. This approach will allow the opportunity to connect theoretical results with practical applications in real-world markets.

Maimoona Irfan

Advisor: Leslie Cheng

Market Disruptions, Climate Risk, and Distributional Consequences: A Mathematical Finance Approach

The fallouts of financial crises have always had disproportionate fallouts, with the brunt falling on the communities least equipped to absorb them. My research this summer aims to examine how mathematical tools can be utilized to visualize and address financial risks that disproportionately affect vulnerable populations.

Specifically, I am interested in exploring two interconnected themes. The first concerns market microstructure and liquidity risk: how do sudden disruptions, flash crashes, volatility spikes, or cascading illiquidity propagate through markets, and who is most affected? Low-income investors, pension-dependent retirees, and under-resourced institutions often lack the hedging mechanisms available to large players, making them uniquely exposed. I hope to model these dynamics using stochastic processes and, where applicable, partial differential equations, analyzing how shocks travel and where interventions might be most effective.

The second theme involves derivatives pricing under macro stress particularly in the context of the energy transition. Carbon markets and renewable energy assets are rapidly growing in importance, yet their pricing models remain underdeveloped relative to traditional asset classes. Mispriced climate-linked instruments can distort capital flows away from the communities and regions most vulnerable to environmental risk. I aim to examine where classical models like Black-Scholes break down under these conditions and explore more robust alternatives.

Lucia Moscola

Advisor: Leslie Cheng

Effects of Sports Gambling Legalization on Traffic Fatalities

Several U.S. states have legalized online sports gambling since a 2018 ruling by the U.S. Supreme Court. Sports gambling has since rapidly gained widespread popularity, with NFL games being a particularly common area of betting. Previous studies have shown that sports gambling exacerbates the rise in domestic violence incidences after unexpected NFL team losses. This suggests that the legalization of sports gambling increases the magnitude of emotional cues caused by these games. Other research has shown that traffic fatalities rise significantly after close wins by NFL teams. In my research, I will investigate whether the legalization of sports gambling has amplified the link between the outcomes of NFL games and the rate of traffic fatalities in associated areas.

Ritika Nuti

Advisor: Chaegeun Song

Play Differences: WNBA vs NBA

For this research project, we want to answer the question of what factors explain higher assist rates in the WNBA vs. the NBA. We will solve this question by doing a statistical analysis. The WNBA historically has had higher assist percentages and focuses more on the team, while the NBA elevates single players. We will use data from the 2024-2025 seasons for both leagues, choosing five teams from each league, and use regression and other testing methods. We will visualize the data in different ways and see what patterns and conclusions we can draw from the data we used. We will use predictors like player height, pace, 3-point percentage, shooting percentage, team efficiency, free throws, and turnovers to see which may have a significant impact on assists. We will also compare how this has changed over time, also examining assist rates and play style from 2000 for all the teams. We also want to see if it is possible to differentiate a WNBA team from an NBA team, looking just at the data. We will also use this data to examine the disparities in data between the WNBA and the NBA and why these exist.

Our data is from Basketball Reference for both of the leagues, and we will do our analysis in R. A potential limitation for our project is that WNBA data is more limited than NBA data, so this may cause an issue in the amount of data we pull from the league. We will also need to standardize the data in some way, as the WNBA has shorter seasons, shorter games, fewer players, and less league history. 

Minahil Saleem

Advisor: Leslie Cheng

Evaluating the Structural Resilience of Binomial Derivative Pricing: An Empirical Simulation Study on the KSE-100 Index

Discrete-time option pricing models are widely used in financial mathematics, yet their performance in high-volatility emerging markets remains underexplored. This paper asks how well the Binomial Option Pricing Model (BOPM) holds up when applied to the Pakistan Stock Exchange (PSX) — a market characterized by political instability, currency stress, and irregular liquidity. We hypothesize that standard binomial pricing, under these conditions, will produce systematic deviations from theoretical benchmarks, and that these deviations will grow more pronounced during periods of macroeconomic stress. Using historical daily data from the KSE-100 index and KIBOR benchmark interest rates, we implement the Cox-Ross-Rubinstein (CRR) framework in Python and benchmark results against Black-Scholes outputs to evaluate model behavior across multiple distinct market regimes. Understanding where and why these models break down has practical implications for derivative pricing in emerging markets more broadly, and contributes to a growing body of literature questioning the universality of models built on assumptions of stable, liquid markets. Future directions may include extending the analysis to alternative models better suited to high-noise environments.

Arlene Ulloa

Advisor: Selvi Kara

Lucky and Displacement Statistics of 3-Stirling Permutations

This research explores parking functions and Stirling permutations, two closely related combinatorial structures. Stirling permutations can be viewed as a special family of parking functions, where each number appears twice and satisfies certain ordering conditions. Recent work has focused on two statistics associated with these objects. Lucky cars, which park in their preferred spots, and displacement which measures how far a car parks from its preferred spot. This project investigates 3-Stirling permutations, which extend ordinary Stirling permutations by allowing each number to appear three times rather than twice.

The primary goal of this project is to understand what happens to the lucky-car and displacement statistics when we move from ordinary Stirling permutations to 3-Stirling permutations. In particular, this research examines which properties and counting results continue to hold, which new patterns emerge, and how these statistics can be used to better understand the structure of 3-Stirling permutations. Questions involving lucky polynomials, admissible lucky sets, displacement compositions, and extremely lucky and extremely unlucky permutations will also be explored.

By extending recent results on ordinary Stirling permutations, this project aims to develop a better understanding of 3-Stirling permutations and their relationship to parking functions.

Yor Daniel Chol Wai

Advisor: Leslie Cheng

 Global Dollar Pressure Index (GDPI).

This paper proposes a new economic index called Global Dollar Pressure Index (GDPI). The GDPI is a composite index that intends to measure and monitor the relative pressure of the US dollar on ten countries, which are a good combination of developed and developing nations. We develop the GDPI by taking the exchange rate, inflation differential, GDP, and Purchasing Power Parity into consideration as primary factors. The data used for this purpose will be gathered from FRED, World Bank, IMF, and Yahoo Finance. The data will be obtained monthly from January 2000 to December 2024. The methodology used includes principal component analysis, K-means clustering, vector autoregression, and rolling Z score standardization. This paper attempts to show that GDPI captures the dollar pressure among different countries and reacts to known economic shocks. Moreover, the methodology is scalable because the index can incorporate more countries if the selected ten countries work well.

Purity Wanjiku

Advisor: Leslie Cheng

A DIFFERENCE-IN-DIFFERENCES EVALUATION OF KENYA’S FREE MATERNITY SERVICES POLICY

My summer research will examine the impact of Kenya’s Free Maternity Services Policy (locally known as Linda Mama), introduced in June of 2013, on the utilization of facility-based delivery services and the reduction of socioeconomic inequities in maternal healthcare utilization. In many developing countries, particularly in Sub-Saharan Africa, poor maternal health outcomes remain a significant public health concern. One of the major determinants of maternal and neonatal outcomes is whether the childbirth takes place in a health facility with skilled birth attendants.

Using data from the Demographics and Health Survey (DHS) conducted in 2008, 2014 and 2022, we will apply a Difference-in-Differences framework and logistic regression approaches to compare changes in delivery outcome between poor and non-poor women before and after the policy’s implementation. We will also examine the probabilistic foundations of causal inference, including concepts such as expected value, variance, estimators, bias, and the use of DHS weights.

We hypothesize that the policy led to a significant increase in facility-based deliveries over time, with poor women experiencing the greatest gains. If supported by the data, these findings would suggest that the policy improved access to maternal healthcare and reduced socioeconomic disparities in healthcare utilization.

Kaizhe (Kath) Cui & Wenyi (Eva) Li

Advisor: Xuemei Cheng

Sputtering Deposition of Centrosymmetric Magnet Films 

Recent studies have identified Gd₂PdSi₃ and GdRu₂Si₂ as centrosymmetric materials that can host nanometer-sized magnetic skyrmions, making them promising candidates for building skyrmion-based qubits [1]. While previous work has focused on bulk single crystals, thin films are needed for device fabrication. 

In this project, Gd₂PdSi₃ and GdRu₂Si₂ thin films will be grown by magnetron sputtering. Sputtering conditions, such as substrates, Ar gas flow, deposition temperature, deposition pressure and deposition time, will be optimized to achieve high-quality epitaxial films. X-ray reflectivity (XRR) measurements will be used to characterize the film thickness and interface quality, while vibrating sample magnetometry (VSM) measurements will be used to investigate the magnetic properties of the films, including hysteresis loops and magnetic anisotropy.

The goal of this work is to establish a procedure for growing these skyrmion-host materials as thin films and to investigate their structural and magnetic properties. The results will shed light on interface magnetism and skyrmion formation mechanism in centrosymmetric magnetic thin films.

Reference:

[1] C. Psaroudaki and C. Panagopoulos, Skyrmion Qubits: A New Class of Quantum Logic Elements Based on Nanoscale Magnetization, Phys. Rev. Lett. 127, 067201 (2021).

Beatriz da Silva & Alyssa Ya

Advisor: Michael Noel

Optimizing laser lock-in stabilization for excitation of Rydberg atoms

We use lasers and magnetic fields to cool and trap Rubidium atoms. These atoms are excited to Rydberg states using additional lasers. The frequency of the laser used for the final excitation step must be precisely controlled in order to select the desired state among a dense array of states. We have built a circuit in combination with a Fabry-Perot to control the frequency of the laser. This summer, we will optimize the design of the circuit and explore different configurations of the cavity to create the most stable system possible.

Marie-Claire Davenport

Advisor: David Schaffner

Title: Just Say Plasma 2, Electric Boogaloo 

Plasma is a state of matter which occurs within space after gas has been heated and ionized. It is defined as a physical system whose intrinsic properties are governed by collective interactions of large ensembles of free charged particles (NSF Basic Plasma Science and Engineering Website). Plasma doesn’t often present itself on Earth due to there being too little energy and higher densities, however it does occur within specific conditions (ie. as lightning and as neon signs). According to the NSF website for plasma physics, plasma potentially governs 99.9% of the visible universe. However, astronomers will refer to astronomical fluids as “gasses,” regardless of if the matter behaves more like plasma. My project aims to quantitatively determine the mass and volume that plasma takes up within the Milky Way, and to potentially serve as outreach material, advising for astronomers and physicists to recognize astrophysical gasses and plasma as distinct states of matter, and to refer to them as such.

Adelaide Grace

Advisor: Asja Radja

The Quantification of Drosera capensis Movements while Trapping Prey-like Stimulus

Drosera capensis, commonly known as Cape sundew, is a carnivorous plant that traps its prey by physically curling or rolling around it. This process is initiated and carried out by tentacles that secrete mucilage, a sweet-smelling sticky substance containing digestive enzymes, to entice, capture, and metabolize its prey. However, this biological process is unexplored and poorly understood. The Radja lab will combine data analysis of the plant to create a quantitative model describing the prey trapping motion of Drosera capensis. This will be done using various sizes of prey-like stimulus made from milk powder and agar agar powder to stimulate these thigmotactic movements (an organism’s response to physical contact). Timelapse photography in combination with MATLAB coding will compute translational velocity, angular velocity, front velocity, and curvature. In addition, confocal microscopy will show the cellular changes during this process. Previous work on this project has shown that the curling is fastest initially and slows as trapping progresses. Additionally, a larger spatial spread of food is shown to catalyze a greater velocity. Microscopically, the cells of the abaxial (lower) side of the leaf expand while closing, while the adaxial (upper) cells remain. My experiments conducted this summer will either further support or challenge these initial findings. Understanding these biological systems can help further advance technology, specifically with the creation of soft robots, and can help create prognostic models for plants.

Oksana Hryvnak

Advisor: Asja Radja

Investigating Water Flow Around Dried and 3D Printed Octocorals

Soft coral (aka octocorals) and hard corals are reef-building colonial animals. Climate change has caused a sharp decline in hard corals, but octocorals have not only remained resilient, but have even expanded into new areas. This project will contribute to the investigation of octocoral resilience by analyzing how they manipulate local water flow. We will study how various factors (including but not limited to porosity, shape, and flow speed) influence the flow of water around octocorals on the scale of branches and colonies. This will be investigated with a flow tank, dried and simplified 3D printed coral samples of varying porosity, size, and shape, and an analysis method called particle image velocimetry (PIV). PIV uses 1-20 micrometer buoyant beads as tracers to determine the speed and direction of water currents. Video data will be analyzed in MATLAB-based PIVlab and Python. We hypothesize that porosity and flow speed will have the strongest impact on flow and leakiness, or the tendency of the coral to act as an impermeable "wall" that particles must go around or a "rake" such that particles can pass between gaps in the coral. It is predicted that at low speeds or porosities water will go around the coral (solid wall), and for high speeds/porosities water will go between the coral (leaky rake). We also hypothesize that the 3D printed coral will behave similarly to comparable dried specimens. Using 3D printed corals allows measurements of porosity or size to be uniform as opposed to the variability of organic coral structures. If 3D coral behaves similar to real corals, it can serve as a generalizable model where more variables can be controlled for. By learning how water flow changes around octocorals, we hope to gain insight into the mechanism of their success to better approach conservation efforts for coral reefs.

Sophie Lewis

Advisor: Michael Schulz

Applications of the AdS/CFT Correspondence to Lifshitz Field Theories

Holography, in quantum gravity, is the idea that a gravitational theory, in any region of space-time, has an equivalent description as a nongravitational theory living on the boundary of that region. Although the conceptual idea of holography had been around for a while, a richer and more quantitative realization of holography came in 1997 and is known as the AdS/CFT correspondence. In this correspondence, there is a complete equivalence between a gravity theory (type IIB string theory on five-dimensional anti-de Sitter space-time) and a nongravitational theory in one fewer dimension space (four-dimensional maximally supersymmetric Yang-Mills theory living on the boundary of anti-de Sitter spacetime). Since the original formulation of the AdS/CFT correspondence, we have come to understand holography as a general tool relating gravity theories to other non-gravitational field theories. The AdS/CFT correspondence is of great practical computational utility, since it is a weak/strong coupling duality. When the field theory is “hard” (strongly coupled), the gravity theory is “easy” (weakly coupled), and vice versa. Thus, we can gain insight into strongly coupled field theories in a variety of subfields of physics (for example, strongly coupled electrons in condensed matter physics and the quark-gluon plasma in particle physics) which would typically be accessible only via numerical computation, by performing explicit and tedious paper and pencil computations in the dual gravity theory. Even when the exact dual gravity theory for the field theory of interest is not known, often there is a very similar field theory whose gravity dual is readily understood, and we can gain insight into the actual field theory of interest from a detailed study of its closely related cousin. In the proposed summer research project, we will reproduce the basic computations of the original AdS/CFT correspondence and then apply these tools to compute correlation functions in condensed matter field theories for systems whose phase transitions are described by Lifshitz-type multicritical points. Such theories arise in the description of magnetic materials and systems with strongly coupled electrons.

Lisa Li

Advisor: Don Fahey

Optimizing Spatial Resolution and Sensitivity in a Quantum Diamond Microscope through a Custom Objective and Pulsed ODMR

Quantum Diamond Microscopes (QDM) utilize nitrogen vacancy (NV) centers for magnetic field imaging, yet the performance of our QDM is limited by a standard objective with low numerical aperture (NA) and continuous-wave (CW) spectroscopy techniques that hinder imaging sensitivity. Our goal this summer is to implement two separate upgrades to improve our QDM. First, we aim to design, assemble, and test a custom, non-magnetic microscope objective specifically optimized for imaging through diamond to achieve an aberration-free imaging and greater light collection via a larger NA. Second, we will transition the QDM from using CW to pulsed techniques by setting up a pulsed LED to perform more precise Ramsey magnetic imaging and measurements with our photodetector and camera. We will update our control software to incorporate the LED current driver, photodetector, and low-frame-rate camera, and test whether the LED intensity is sufficient to repolarize the NV centers in pulsed operation. Together, these improvements will allow us to achieve better sensitivity on our QDM for rock magnetism experiments, and enable AC magnetic imaging for future meta-material studies.

Sara Mandour

Advisor: Don Fahey

Enabling Ramsey QDM Magnetometry via a PCB Microwave Resonator and Helmholtz Coils

Ramsey Magnetometry is used in quantum systems like diamond nitrogen-vacancy (NV) centers to detect magnetic fields due to its exceptional magnetic sensitivity. Our research upgrades our existing quantum diamond microscope (QDM) with new tools that enhance sensitivity and enable new modalities. We will design, simulate, build, and test 3-axis Helmholtz coils that will provide a highly uniform, controllable bias magnetic field for tuning the NV centers in our diamond samples. Additionally, we desire the ability to apply intense and uniform microwave fields (near 2.87 GHz) that are necessary for Ramsey magnetometry across a 2x2 mm region of the diamond. We will adapt a previous design for a PCB microwave antenna to deliver large, homogeneous Rabi frequencies across field-of-view of our QDM. If successful, we will modify the design to produce circularly polarized fields that selectively drive single spin transitions. When implemented, these tools will significantly improve our QDM capabilities and usability for a range of wide-field imaging experiments targeting geological, biological, and metamaterial samples. 

Alexandra Morales

Advisor: Don Fahey

Design and Implementation of an All-Optical NV-Center Magnetometer

Nitrogen vacancy centers in diamonds exhibit properties that enable magnetometry in wide ranging conditions. By shining a green laser light onto an NV center, the emitted photoluminescence (PL) allows for the characterization of a measurement of the magnetic field. The most common protocols for NV sensing involve the use of microwaves (whether continuous-wave or pulsed) and are rather large in setup. A newer method, known as all-optical photoluminescence (AO-PL), is potentially beneficial because it does not require the use of microwave fields. This simplifies the design and makes it more robust to extreme environments.

This research investigates the parameters necessary to optimally sense magnetic fields, and then implements the parameters in a test assembly. We will first replicate the proof-of-concept results and optimize the sensitivity on the existing apparatus. In particular, we will determine the optimal laser intensity and lock-in detection parameters. We will then construct a purpose-built apparatus the includes a temperature-controlled stage and repeat the measurements for a range of temperatures. We will assess the performance metrics (e.g. sensitivity, dynamic range, bandwidth) and the potential for deploying this sensor in extreme environments including space and plasmas.

Laura Mujica

Advisor: David Schaffner

Investigating the magnetic field and density variations of the solar wind’s interactions with the Moon

Interactions between the Moon and the solar wind, especially the mechanism that creates the lunar wake, a plasma region downstream the moon, have been studied for decades. How plasma interacts with small objects and the behaviors that occur in this scenario are being simulated and studied in laboratory experiments using the Bryn Mawr Experiment (BMX) by observing how magnetic field and density of plasma at speeds up to 100 kilometers per second changes due to a small object blocking its flow.

To understand this mechanism and what causes the real-world situation in which large plumes of plasma interact with a relatively small object, data, statistical, computational, and graphical analysis will be done on recorded measurements of magnetic field and density of plasma around the moon from the two-spacecraft ARTEMIS mission. 

This research focuses on fluctuations of magnetic field, density, and other parameters of the solar wind as it interacts and passes through the Moon to compare data and findings to those of laboratory simulations using plasma and an obstacle at a much smaller scale in the Bryn Mawr Experiment.

Sofia Sergi

Advisor: David Schaffner

Using a HeNe Heterodyne Interferometer to Analyze Plasma Density

Plasma is an ionized state of matter characterized by its collective, non-local behavior and its interactions with electric and magnetic fields. Interferometry is often used to measure line-integrated plasma density. In this project, we will be measuring the density of fully ionized hydrogen plasma created in the Bryn Mawr Experiment (BMX). The BMX consists of a copper gun source in which hydrogen gas is ionized, a diagnostics section, and a stainless-steel dump section. Our goal this summer is to construct a heterodyne interferometer with unequal path length. Using an acousto-optic modulator, we will split the beam of a red HeNe laser into two: a reference beam, which will travel through air, and a plasma beam, which will travel through a plasma chamber located in the diagnostics section of the BMX. After the two beams travel through their configured paths, we can superimpose them to reveal a phase change, which can be used to calculate the electron density of the plasma. Because we know what the density should be, this project will give us broader insight into what methods of plasma density measurement are most efficient, convenient, and accurate.

Adara Alexander

Advisor: Thomas Le

“I can handle it”. The Strong Black woman schema and the link between emotional invalidation and disordered eating amongst black women ages 18-26

Black women and stress have become synonymous. Black women exist in a precarious position where we are both undervalued and invalidated while tasked with being leaders and uplifting those around us, forcing us to put the needs of others before ourselves. This constant drive to present a brave face and to mask our emotions at the cost of our health is referred to as the Black Woman schema, which stems from the historic racism, sexism, and misogynoir that Black women endure. This schema, categorized by emotional suppression, self-reliance, and resilience at the cost of one's own health, is deleterious to Black women’s mental health. Furthermore, emotional invalidation, categorized as the dismissal and/or minimisation of one's feelings, is experienced by eight out of ten Black women and has been associated with disordered eating behaviors. Emotional invalidation, in tandem with the Strong Black Woman schema, creates a cycle where Black women have little room for rest, which can lead to skipping meals. My proposed project investigates if internalisation of the Strong Black Woman schema mediates the relationship between emotional invalidation and disordered eating amongst black women ages 18-26. This study will provide much-needed research into the harm of emotional invalidation and the Strong Black Woman stereotype that are pushed on Black women. Research questions will be assessed using regression and mediation analysis on survey data collected from a sample of African American women.

Sungwon Bae

Advisor: Thomas Le

Bicultural Integration as a Protective Factor against Muscularity-Oriented Disordered Eating among Asian American Men: Acculturative Stress & Affirmative Socialization as Mediators

Eating disorder research often overlooks how pathologies manifest in men of color. For Asian American men, body dissatisfaction frequently presents as muscularity-oriented disordered eating, involving a drive for a hyper-muscular physique based on rigid dieting and excessive exercise. This struggle is rooted in the intersectionality of Asian identity and gender norms, conflicting with both Eurocentric beauty standards and racialized ideals of masculinity.

Bicultural integration is the ability to reconcile Asian heritage with American identity. For Asian Americans, this skill may serve as a protective buffer. By fostering a fluid approach to beauty, individuals can adopt realistic standards that incorporate their racial identity rather than internalizing unrealistic Western ideals. Furthermore, bicultural competence facilitates diverse social networks, providing healthier coping mechanisms and social support.

This study examines two primary mediators. First, biculturalism may reduce acculturative stress which is often associated with disordered eating. Second, bicultural integration promotes diverse social networks, increasing affirmative socialization. I hypothesize that greater affirmative socialization fosters psychological safety which subsequently reduces disordered eating. By exploring how an integration protects against the intersection of racism and harmful masculine ideals, this research addresses a critical gap in clinical literature. Ultimately, this work offers nuanced insights into Asian Americans’ well-being, with implications for culturally responsive prevention and treatment.

Ellie Carlin

Advisor: Cora Mukerji

What are the unique effects of harshness and predictability in childhood vs adolescence on cognitive flexibility?

Existing research has found associations between adverse childhood experiences (ACEs), such as abuse, neglect, and unstable households, and negative consequences later in life, such as increased risk for difficulties in executive functioning in adulthood (Lund et al., 2020). Cognitive flexibility (CF) can be defined as one’s mental ability to shift between perspectives, tasks, and concepts, to adapt one’s patterns of thinking to new contexts. Studies suggest that environmental experiences that occur between ages 7-9 may be particularly impactful for developing CF skills (Anderson, 2002), and CF becomes relatively mature and reaches adult levels of performance around age 12 (Cepeda et al., 2001; Crone et al., 2006). Recent research on dimensions of ACEs suggest that while environmental harshness, i.e., conditions that increase morbidity and mortality, may constrain cognitive flexibility, exposure to unpredictable environments in childhood and/or adolescence may contribute to increased cognitive flexibility in some circumstances (Fields et al., 2021; Mittal et al., 2015; Pope-Caldwell et al., 2024; Niemiec et al., 2026)Notably, comparatively little work to date has explored the effects of exposure to predictable childhood environments on cognitive flexibility.

Using data from a nationally-representative sample of ~400 adults, I will test the relationship between experiences of predictability during childhood vs adolescence and participants’ cognitive flexibility. Through a multiple regression model, I will evaluate associations between participants’ self-reported caregiver predictability during childhood (ages 6-12) and adolescence (ages 13-18) from the Predictability, Opportunity, and Safety Dimensions of Environmental Variability Questionnaire (POS-DEV), environmental harshness (as indexed by childhood socioeconomic status), and their cognitive set-shifting performance on the Dimensional Change Card Sort (DCCS) task. I hypothesize that predictability during childhood will explain more variation in adult cognitive flexibility performance than predictability during adolescence; however, this effect could be moderated by harshness, such that exposure to high levels of harshness will weaken the association between childhood unpredictability and CF. This proposed study has important implications for caregiver predictability during developmental periods and potential connections to adult executive functioning.

Tanu Erdembileg

Advisor: Yeon Soon Shin

Investigating how curiosity predicts exploration in varying environments

Curiosity is an intrinsic drive for information even in the absence of immediate rewards (Kidd & Hayden, 2016; Gottlieb & Oudeyer, 2018). Previous studies have shown that states of higher curiosity enhance memory for incidental information encountered compared to lower curiosity (Murphy et al., 2021; Cen et al., 2021), suggesting that curiosity motivates learning and facilitates better retention of information gained from exploration. While previous work shows that curiosity correlates with greater spatial exploration, with individual differences in trait curiosity modulating this relationship during navigation (Cen et al., 2024), it remains unclear which aspects of the environment increase exploration and which dimensions of curiosity drive this behavior. One such dimension is stress tolerance, which reflects one’s tendency to handle anxiety when confronted with novelty (Kashdan et al., 2020), and may help determine how willing an individual is to venture into unfamiliar new environments. 

Our research aims to investigate how curiosity relates to the underlying decision-making process that guides exploration across environments with differing diversity (the variety of objects) and scarcity (the number of objects). To determine whether curiosity modulates which rooms participants choose to prioritize during navigation, we use a two-dimensional grid navigation and memory retrieval task, where participants navigate a simulated zoo with four rooms of varying diversity and scarcity to locate objects. They are then asked to identify whether they can recall certain objects and their corresponding location. We predict that higher levels of curiosity will lead to more exploration towards rooms with greater diversity of objects, leading to more deviation from reward-optimal paths and revisits, and less exploration in scarce rooms with fewer objects. Our preliminary research has shown that Maximum Occupancy Principle (MOP) agents, which prioritize maximizing the amount of space covered in the environment, deviate more frequently from reward-optimal paths than epsilon-greedy Q-learning agents. This project may further explore how factors such as stress tolerance relate to variability in curiosity, spatial exploration, and navigation.

Ciara Grimes

Advisor: Anjali Thapar

Application of Eye-Tracking Measures to Isolate Components of Cognitive Load

Previous research has demonstrated that eye-tracking measures such as pupil dilation and gaze are sensitive to cognitive load (Liu et al., 2022). However, in this research, cognitive load is typically treated as a singular construct. In contrast, theoretical work has demonstrated that cognitive load is a multidimensional construct. Two core components of cognitive load are working memory, which is required to maintain and manipulate information, and attentional control, which allows for the focus on relevant stimuli. Commonly studied eye-tracking measures associated with cognitive workload include pupil measures, fixation duration and count, saccades, and blink rate (Xiaofu et al., 2025). The goal of this study is to determine whether specific eye-tracking measures are differentially sensitive to working-memory load versus attentional control. To address this question, participants will complete two cognitive tasks: an n-back to manipulate working-memory updating and a mental arithmetic task to manipulate attentional control demands. Based on previous findings, I hypothesize that pupil dilation, blink rate suppression, and reduction in microsaccade rate will be more sensitive to working-memory load, suggesting increased cognitive effort during updating processes (Krejtz et al., 2018; Rondeel et al., 2015; Xiaofu et al., 2025). I also hypothesize that fixation duration and fixation count will be more sensitive to attentional control (Xiaofu et al., 2025). The results of this study will provide valuable insight into how specific components of cognitive load relate to eye-tracking measures.

Sarah Haskell

Advisor: Ariana Orvell

Assessing Frustration Inducement in Children

Emotional regulation skills are critical to socioemotional wellbeing later in life (Chervonsky & Hunt, 2019; Cloitre et al., 2019; Polizzi & Lynn, 2021), and children rely on adults to provide structure and support as these skills develop in early childhood (Gee & Cohodes, 2021). The majority of current literature uses correlational methods to assess parents’ use of multiple coping strategies in combination, leaving ambiguity as to the efficacy of parental support using specific emotional regulation techniques. Here we focus on emotional validation, in which one affirms understanding of another’s emotion, and positive reappraisal, the act of mentally reframing a negative event to see the “bright side” (Sahi et al., 2023). These techniques are assumed to be helpful (Jeon & Park, 2024; Willner et al., 2022), but whose efficacy has not been directly compared in experimental studies where parents provide regulatory support for their children. The Nurturing Emotion Skills Together (NEST) Study aims to evaluate whether cueing parents to support their children using emotional validation vs. positive reappraisal effectively reduces child frustration, compared to a neutral control. Parents of children aged five to seven will learn about emotional validation, positive reappraisal, or sleep (control condition). Children will then be given a challenging puzzle designed to induce frustration, and parents will be prompted to help their children emotionally regulate using condition-specific phrases. A pilot study will be run prior to the full study to validate that the puzzle task induces moderate frustration in this age range, has similar difficulty across demographics, and has developmentally appropriate difficulty. Participants will be asked to complete four puzzles—two designed to be easy and two designed to be overly difficult—randomly selected from a repository of ten similar puzzles, each within a two-minute time period (Nolen-Hoeksema et al., 1995; Gentzler et al., 2013). Parents will be instructed to not interfere during the pilot task. Children will report their baseline frustration level (pre-task) on a Likert scale, again after each puzzle, and retrospectively at the end of the tasks. Parents will also report their perception of their child’s frustration during the hard/easy puzzles compared to everyday frustrating stressors. To ensure that frustration is induced by the difficulty level of the puzzle itself, children and parents will be asked to assess ease of game navigation. These data, along with the rate of successful completion of these puzzles will be used to select final stimuli for the main study. 

Wilhelmina Lawerh-Lawerh

Advisor: Cora Mukerji

Associations Between the Developmental Timing of Caregiver Predictability and Perceived Stress in Adulthood

Positive childhood experiences (PCEs) are supportive relationships and environments associated with healthy development and well-being across childhood. A growing body of research suggests that cumulative PCEs may be associated with lower perceived stress and better mental health outcomes in adulthood, even among individuals exposed to adverse childhood experiences (ACEs) (Bethell et al., 2019). Recent research on childhood adversity suggests that different dimensions of ACEs may influence development in distinct ways (McLaughlin et al., 2021); moreover, the developmental timing of these experiences may shape whether—and how much––development is affected (Gee & Casey, 2015; Cohodes et al., 2023). However, comparatively little is known about how different dimensions of PCEs, such as predictability, opportunity, and safety, and their timing impact mental health and stress (Lopez et al., in press). Caregiver predictability, in particular, has received growing attention in research as it has been linked to resilience, stress regulation, and long-term psychological functioning (Davis & Glynn, 2024; Glynn, 2025). Developmental timing may be particularly important when examining caregiver predictability: childhood and adolescence are distinct periods of neurodevelopment and psychosocial change, and prior research suggests that experiences occurring during these periods may have different implications for future functioning (Gee & Casey, 2015; Sisk & Gee, 2022). 

Using questionnaire data from a nationally-representative sample of adults (n=401), I will investigate whether caregiver predictability during childhood and/or caregiver predictability during adolescence (Predictability, Opportunity, and Safety Dimensions of Environmental Variability Retrospective Scale; Lopez et al.., in press) are uniquely associated with adult perceived stress (Perceived Stress Scale-10; Cohen et al., 1983), while controlling for cumulative ACE exposure (Philadelphia ACEs; Cronholm et al., 2015) and childhood socioeconomic status (SES). I hypothesize that greater caregiver predictability during both childhood and adolescence will be associated with lower levels of perceived stress in adulthood. Additionally, I expect that caregiver predictability during childhood may show a stronger association with perceived stress than caregiver predictability during adolescence, as consistent with findings from Cohodes et al. (2023). Findings from this study can help inform future interventions and prevention efforts aimed at increasing predictable caregiving environments during development as a means of promoting long-term psychological well-being.

Lianna Pan

Advisor: Arianna Orvell

Equipping Parent-Child Interactions with Emotional Regulation Science 

Emotional regulation (ER) is the ability to modify emotions to accomplish one’s goals (Thompson, 1994). Young children skilled in ER show many dimensions of well-being across the lifespan, including early school adjustment and social competence (Adynski et al., 2024). Parents play a crucial role in shaping children’s intrinsic ER development (Gee & Cohodes, 2021; Rutherford et al., 2015). 

Notably, however, little empirical research educates parents about the science of ER and assesses if doing so positively shapes children’s ER development. Currently existing experimental designs that teach parents ER strategies to reshape parental-child interactions are time-intensive, limited in accessibility, and narrow in demographic reach (Burkhardt et al., 2024; England-Mason & Gonzalez, 2020).

Our three-phase study teaches parents evidence-based strategies known to facilitate ER using short, accessible interventions. The focus groups phase aims to understand how parents attempt to help children regulate their emotions, thoughts, and behavior in daily life. Phase two is a descriptive study that tracks parents’ use of ER with their children in real time using daily diary methodology. Phase three implements a 20-minute intervention to teach parents ER strategies informed by phases 1 and 2’s findings, as well as extant literature. Parents will record their progress using daily diary surveys for two weeks, reporting strategy use, self-efficacy, and child behavior.

We will leverage principles from the community-participatory research framework to complete phase one’s pilot work. More specifically, through focus group interviews, we will first identify parents’ ER frequency, strategies, and stressors in the Philadelphia area, and then evaluate how they map onto commonly studied taxonomies, such as Gross’s (1998) five-strategy framework. We will also note characteristics that may moderate parental ER competency identified in previous literature, such as mood disorders, developmental stability, SES, and culture.

Sonia Presser

Advisor: Dustin Albert

Predictors of Parent-Child Congruence in Reporting Adolescents’ Internalizing Symptoms

Mental health problems like depression and anxiety, collectively studied as internalizing disorders, demonstrate sharply increased prevalence in adolescence. Adolescence is also characterized by increased motivation to attain autonomy from parents and to establish a distinct identity. For many families, the gap between parents’ understanding of their adolescent’s inner lives and the adolescent’s lived experience widens, which may contribute to even greater risk for mental health problems. 

The multiple informant model is a popular approach to child psychology research in which data about an individual is gathered from multiple sources (e.g. the individual’s self-report, parents, or teachers). This method is used to create a balanced and nuanced representation of individuals. Although each informant is asked the same questions, discrepancies often occur. These discrepancies are not accidental, but can offer us insight into the parent-child relationship and overall patterns in adolescent mental health. Parent-child congruence in the reporting of adolescents’ internalizing symptoms shows how accurate a parent’s perception of their child’s mental health is. Lower parent-child congruence (i.e. discrepancies in perception) is associated with internalizing and externalizing symptoms in children (De Los Reyes et al., 2008). Additionally, parent-child congruence is typically higher when reporting externalizing symptoms as compared to when reporting the less observable internalizing symptoms (De Los Reyes et al., 2015).
We will be analyzing how different characteristics of the parent correlate with internalizing symptom congruence, such as the parent’s mental health, beliefs about parenting, warmth, discipline, and time spent together. These analyses will be conducted on the Parenting Across Cultures dataset, a study that uses the multiple informant model by interviewing parents and their children from ages eight to twenty-three. Its large sample of nearly 1,500 participants across nine countries will increase the external validity of findings in this field. 

Priyanka Regmi

Advisor: Thomas Le

Colorism and its impact on ADHD symptoms among South Asian American Adults: The Roles of Skin Tone Surveillance and Skin Shade Satisfaction

Attention-deficit/hyperactivity disorder is one of the most common neurodevelopmental disorders, characterized by neurodivergent levels of hyperactivity, impulsivity, and inattention (American Psychiatric Association, 2013; Boyle et al., 2011). Although ADHD has been extensively studied in children and increasingly in adults, research on ADHD within racial and ethnic minority populations, particularly South Asian Americans, remains extremely limited.

Limited research has highlighted that race-related trauma and chronic psychosocial stressors can influence ADHD symptom expression (Boyd et al., 2024), suggesting that sociocultural experiences may exacerbate attentional and executive functioning difficulties. Within the South Asian community, colorism, often conceptualized as discriminatory treatment based on having a darker skin tone, is a pervasive stressor that may contribute to these outcomes (Sadek et al., 2026). 

Prior research with Indian women has shown that skin tone surveillance is associated with poorer well-being outcomes, such as higher depression and lower life satisfaction (Prusaczyk & Choma, 2018). Thus, high levels of skin tone surveillance may increase stress, deplete cognitive resources, and interfere with attention and self-regulation, potentially mediating the relationship between colorism and elevated ADHD symptoms.

Similarly, skin shade satisfaction, the degree to which individuals feel positively about their skin tone, may moderate the impact of colorism on ADHD symptom expression. Because stress responses overlap with core features of ADHD, ongoing racialized and colorism experiences may intensify or mimic ADHD symptoms. 

The present study aims to examine 1) the extent to which colorism is associated with Attention deficit hyperactivity disorder (ADHD) among South Asian American adults, 2) the

extent to which skin tone surveillance mediates this association, and 3) the extent to which skin shade satisfaction moderates this association.

Bea Riley

Advisor: Yeon Soon Shin

The effects of MDMA, Psilocybin, and Methylphenidate on Moral Inference

The ability to accurately judge the moral character of others is important for successful social interactions, but the mechanisms underlying this ability are not yet fully understood. Research using the Moral Inference Task (MIT), has participants predict the decisions of people who differ in their willingness to harm others for financial gain. The varied moral impressions are updated through an asymmetric Bayesian mechanism. Specifically, beliefs about morally bad people are more uncertain and flexible than beliefs about morally good people, which allows for negative impressions to be revised more easily (Siegel et al., 2018). 

People who present with more manipulative or explosive traits typically have a more pessimistic prior expectation of others, and in the study, they presented with reduced accuracy in predicting moral decisions and less trustworthy behavior (Shin et al., 2024). Together, these findings suggest that moral inference plays an important role in adaptive social functioning. However, it is still unknown whether pharmacological interventions can alter the processes in which people form and update beliefs about other people’s moral character. This project examines the effects of MDMA, psilocybin, and methylphenidate (active placebo) on moral inference in healthy adults.

Given the evidence that psilocybin boosts flexible thinking (Kuypers, 2018), and that MDMA enhances emotional empathy and alters emotional processing (Hysek et al., 2013, Schmid & Bershad, 2024). I hypothesize that psilocybin will change how prior and evidence is integrated in belief updating, while MDMA’s effect on belief updating will depend on one’s sensitivity to financial gain versus harm inflicted on another person. I further predict that these effects will persist beyond the acute drug effect stage and remain clear at the one-month following up. If found, these findings would suggest that serotonergic and psychoactive compounds can permanently alter moral reasoning; this could impact treatments for disorders involving impaired social functioning such as schizophrenia (Arnovitz et al., 2022).

Rebecca Rothstein-Epp

Advisor: Anjali Thapar

Title: Olfactory Identification Performance as a Predictor of Age-Related Changes in P300 Response and Cognitive Functioning 

Olfactory dysfunction is commonly found in older populations and has been linked to cognitive decline and neurodegenerative disease. One potential explanation attributes the dysfunction to a reduction in the cortical volume of olfactory nervous structures that are often implicated in neurodegeneration. Furthermore, the ability to detect odorants also predicts performance on neuropsychological tests (Uchida et al., 2020). These findings suggest olfactory performance may serve as a biomarker for age-related cognitive changes. In addition to behavioral measures, age-related changes in neural processing can be assessed using event-related potentials (ERPs) derived from electroencephalography (EEG). The P300 component, which is commonly elicited by an auditory oddball task, is thought to reflect attentional allocation and stimulus evaluation processes. Prior studies have observed age-related changes in the amplitude and latency of the P300 component (Criel et al., 2023). The present study investigates whether olfactory identification performance, as measured by the Sensonics’ Smell Identification Test, predicts age-related differences in the P300 component and cognitive functioning. It is hypothesized that poorer olfactory identification performance will be associated with longer P300 latency, smaller P300 amplitude, and lower neuropsychological test scores. The results of this study may have applications in detecting early markers of age-related changes in cognition. 

Fiona Shen

Advisor: Yeon Soon Shin

Altering Belief Updating in Moral Inference

Moral inference, a process of updating beliefs about others’ character by observing their actions (Siegel et al., 2018), is essential for adaptive social functioning. However, the neural mechanism for belief updating in moral inference is less well known. In belief updating, psychedelics are thought to weaken the brain's top-down processing and enable revision of beliefs (Carhart-Harris & Friston, 2019), fundamentally altering a person's social functioning (Weiss 2021) and moral values (Kähönen et al., 2024; McGovern et al., 2022) and sometimes leading to false insights (McGovern et al., 2024). The present study investigates how psychedelics like psilocybin drive downstream distortions in moral inference by using a Moral Inference Task. This trial-by-trial prediction task tracks how individuals form mental representations of others' moral character by observing agent behaviors. By using a computational inference model, we test how these substance-induced distortions change social prediction accuracy and alter the certainty of subjective impressions. We hypothesize that psychedelics like psilocybin will alter the capacity to learn from social experiences, and thereby the ability to accurately update their impressions of harmful behavior and meta-cognitive certainty that reflects accuracy of beliefs. Ultimately, this research offers empirical evidence for how psychedelics change belief updating in moral inference, bridging the gap between psychopharmacology and moral psychology.

Mia-Fay Swanson

Advisor: Ariana Orvell

The Role of Gender in Rituals' Effects on Children's Self-Control

Children frequently encounter situations that require them to practice self-control—the prioritization of distant, abstract goals over more immediate, impulse-driven goals (Fujita, 2011). Self-control, however, frequently necessitates significant cognitive effort, which can pose a challenge for young children due to their developing cognitive capabilities (Lee et al., 2013). 

Existing research on adult self-control suggests that carrying out rituals, or repetitive series of actions performed in a set sequence, redirects cognitive effort onto physical actions and elevates a sense of self-discipline (Tian et al., 2018). It has yet to be studied whether the positive effects of rituals on self-control extend to children, particularly regarding novel rituals that precede a difficult activity. The efficacy of rituals on children’s self-control may also depend on gender, as females often demonstrate higher self-control than males starting from an early age (Chapple et al., 2010; Duckworth et al., 2015; Hosseini-Kamkar & Morton, 2014) due to  socialization practices in which parents monitor and discipline girls for misbehavior more, in adherence with sociocultural expectations (Gottfredson & Hirschi, 1990; Jo & Bouffard, 2014).  Research yields mixed findings on gender-specific behavioral change following self-control interventions. Studies which reported greater change in females suggest their already higher self-control enables them to more successfully execute behavioral interventions and reap benefits (Mansouri et al., 2016; Piquero, 2009). In contrast, studies that find males benefit more from self-control interventions posit that their lower baseline-self-control leaves more room for improvement (Lakes & Hoyt, 2004; Friese et al., 2017). I will examine how gender moderates the effect of novel rituals on children’s self-control and perseverance. Our study instructs children to help a researcher complete a boring peg-turning task. If children become bored, they have the option of switching to a more fun game. Children are randomly assigned to either perform a guided ritual prior to the task or to a control group that performs unscripted movements. We will compare children’s time spent and performance on the boring task hypothesizing that engaging in the ritual will predict longer persistence. Regarding gender differences, we foresee competing hypotheses given prior literature and will conduct simple effects to examine if females or males demonstrate greater persistence in response to the ritual manipulation.

Mahveen Wani

Advisor: Cora Mukerji

Does self-esteem mediate the relationship between positive and adverse childhood experiences and internalizing mental health symptoms?

Adverse Childhood Experiences (ACEs) are associated with a wide range of psychological outcomes, including elevated symptoms of depression (Pantagou and Macbeth, 2022; Elmore and Couch, 2020) and anxiety. (Elmore and Crouch, 2020; Kim et al., 2022) Recent studies also demonstrate a negative relationship between ACEs and self-esteem, which, in turn, contributes to individual differences in psychological distress and internalizing symptoms. (Thilakwardana and Jayasinha, 2025; Annett et al., 2023; Li et al., 2023) This framework positions self-esteem as a critical mediating variable that partially explains the relationship between ACEs and psychological outcomes. Emerging research has also begun to explore the relationship between Positive Childhood Experiences (PCEs) and psychological outcomes, including mental health (Qu et al., 2022; Crandall et al., 2019; Narayan et al., 2019) and self-esteem, demonstrating a positive association. (Kocatürk and Çiçek, 2021) Preliminary evidence also suggests that PCEs are linked to psychological well-being even in individuals with higher ACEs. (Lopez et al., in press; Han et al., 2023; Hinojosa and Hinojosa, 2024) However, existing research has predominantly treated PCEs as unidimensional, leaving the specific multidimensional nature of early positive environments underexplored.

The present study helps address this gap by exploring the effects of childhood safety as a dimension of PCEs using the Predictability, Opportunity, and Safety Dimensions of Environmental Variability (POS-DEV) scale. (Lopez et al., in press). Using questionnaire data from a nationally representative sample of U.S. adults, I propose to model associations among retrospectively-reported childhood adversity (Philadelphia ACEs questionnaire; Cronholm et al., 2015), self-esteem (Rosenberg Self-Esteem Scale; Rosenberg, 1985), and internalizing symptoms in adulthood (PHQ-9 and GAD-7; Kroenke et al., 2001; Spitzer et al., 2006). This study specifically tests whether self-esteem acts as an explanatory variable linking childhood safety to internalizing symptoms in adulthood. I hypothesize that self-esteem will partially explain the relationship between childhood safety and adult internalizing symptoms, such that greater childhood safety will be associated with higher self-esteem, which in turn will be associated with fewer symptoms of anxiety and depression; in contrast, I predict that greater exposure to childhood adversity will be associated with lower self-esteem, which, in turn, will be associated with greater symptoms of anxiety and depression. Findings could help inform efforts to provide trauma-informed care by prioritizing the cultivation of safe relational environments across caregivers, peers, and mentors.

Jadyn Frank

Advisor: Ariana Orvell

Examining associations between child temperament and persistence in a task requiring self-control

Children are often expected to complete boring but necessary tasks that require self-control (e.g., homework). Children vary in their ability to exercise self-control, and self-controlled behavior is associated with differences in temperament, including effortful control and negative affectivity (e.g., irritability; Putnam & Rothbart, 2010; Grennell et. al 2019; Garon & Moore, 2010). However, children’s self-control can also be shaped by self-control strategies, like those that help pull focus away from temptation (Mischel et. al, 1972). Here we examine an unstudied potential self-control strategy for children: rituals—symbolic sets of actions that are structured and repeated—which are theorized to promote self-control by heightening self-discipline. Here, we will assess whether rituals promote self-controlled behavior in children, and whether their efficacy is moderated by individual differences in child temperament. Before participating in the study, parents were asked to complete the Children’s Behavior Questionnaire Very Short Form (CBQ-VSF) to assess their child’s temperament (Putnam & Rothbart, 2010). In a moderated Zoom session, children (N = 128, ages 6-9) were introduced to a 15 minute boring peg-turning task, with the option to take breaks and play a more enjoyable game. Prior to engaging with the tasks, participants were asked to do three movements; in the control group they were random and in the experimental condition they were told to do three specific movements three times. Given prior research (Garon & Moore, 2010), we hypothesize that children with high negative affectivity or low effortful control will spend less time on the boring task and will be more likely to stop the task early. We further hypothesize that children with high negative affectivity or low effortful control will exhibit more persistence in the ritual condition than children with these traits in the control group, based on conceptually similar work (Grennell et. al, 2019). The results of this study will provide insight into how effective rituals may be as a self-control strategy based on child temperament, and can be used to help parents/caregivers decide if rituals may be a useful self-regulation task for their child to engage in before performing a boring, but necessary task.

Michelle Zhao

Advisor: Cora Mukerji

The Association Between Adverse Childhood Experiences and Frontal Alpha Asymmetry Across Threat-Related and Deprivation-Related Dimensions

Adverse Childhood Experiences (ACEs) are harmful events, such as abuse, neglect, or household dysfunction occurring before age 18, are associated with a range of adverse psychological and behavioral outcomes, including increased risk of psychological disorders (Felitti et al., 1998; Hughes et al., 2017). Recent work employing a dimensional approach to studying ACEs has found threat-related experiences (e.g., abuse) and deprivation-related (e.g., neglect) experiences during childhood may have partially distinct effects on the development of brain and behavior (Sheridan & McLaughlin, 2014). While exposure to childhood threat tends to be associated with heightened emotional reactivity, exposure to childhood deprivation is often linked to challenges in cognitive control and regulatory functioning. A convergent line of research using electroencephalography (EEG) suggests that frontal alpha asymmetry (FAA), a neurophysiological metric used to measure the difference in brain activity between the left and right frontal hemispheres, may reflect trait-level differences in people’s motivational tendencies toward approach/avoidance in everyday life (Reznik & Allen, 2018). While several studies have suggested the childhood adversity may shape the development of FAA (Im et al., 2022; McLaughlin et al., 2011), few studies to date have examined unique associations between specific dimensions of childhood adversity and FAA later in life. Using data from the Life Experiences and Outcomes (LEO) study, we will examine associations between estimates of alpha asymmetry derived from resting EEG data and childhood experiences of deprivation and threat assessed using a standardized retrospective questionnaire. Using linear regression, we will test whether threat-related and deprivation-related ACEs show distinct associations with FAA. I hypothesize that (a) both threat-related and deprivation-related ACEs will be associated with greater relative right-frontal activity compared with individuals who did not experience ACEs, (b) this association will be greater in threat-related ACEs than deprivation-related ACEs, given evidence linking high threat sensitivity to greater avoidance motivation and right frontal asymmetry (Heffer & Willoughby, 2020). Understanding how threat-related and deprivation-related ACEs are associated with neural trajectories can help to predict long-term psychological and behavioral outcomes and improve corresponding interventions.