Physics
Maximum-Entropy Exploration Using Intrinsic Rewards
Presenter: Adam Kamoski
Faculty Sponsor: Rahul Kulkarni
School: UMass Boston
Research Area: Physics
Location: Poster Session 1, 10:30 AM - 11:15 AM: Room 165 [D10]

The problem of efficient exploration has been a central focus in reinforcement learning for several decades, yet it remains an open problem to learn policies that sample novel states. A common approach involves using a maximum-entropy objective to encourage broad coverage of the state space. However, standard methods used to estimate entropy can be computationally expensive. Recent work in the average-reward setting has derived an analytical expression for entropy, suggesting a more efficient approach. Building on this result, we determine a reward function that implicitly induces an entropy-maximizing policy. We show that our experimental results agree with analytical predictions, and we demonstrate our framework's general applicability to the efficient exploration problem.

RELATED ABSTRACTS

To Emulate the Stars: Nuclear Fusion as an Energy Production Alternative
Presenter: Nicholas Christopher Giancioppo
Faculty Sponsor: Clarissa Codrington
School: Massachusetts Bay Community College
Research Area: Physics
Location: Poster Session 2, 11:30 AM - 12:15 PM: Campus Center Auditorium [A84]

As the global consumption of electricity increases each year, so too does the demand for alternative sources of energy production. One proposed alternative, nuclear fusion, appears to be a promising avenue in the ongoing pursuit of a healthier, more energy-efficient world. Though still rather early in its development, breakthroughs within recent years have displayed the efficacy of artificial fusion at small scales and bolstered confidence in the eventual production and implementation of large scale nuclear fusion reactors. Not too dissimilar from nuclear fission; fusion utilizes the energy released from atomic activity. Fusion, however, generates several times more energy in proportion to the mass of its fuel than fission. Other strengths of nuclear fusion include its exceptionally low production of waste materials, which are short-lived and not a cause for environmental concern, as well as its ability to operate in a highly controlled environment, ensuring safety. Overall, fusion-based nuclear reactors hold much potential for future means of energy production and hold distinct benefits such as greatly increased efficiency, cleanliness, and stability. This poster will investigate these areas as compared to other major sources of energy, including fission, fossil fuels, and renewables, using data collected by primary and secondary scholarly sources produced by experts in their respective fields. The poster and presentation will serve as an insightful summary of the benefits of nuclear fusion and highlight the importance of continued research and resource investment in its development. 
Orbital Simulation and Satellite Animation
Presenter: Declan Noonan
Faculty Sponsor: Andrew Burkhardt
School: Worcester State University
Research Area: Physics
Location: Poster Session 2, 11:30 AM - 12:15 PM: Room 163 [C16]

I built a Python orbit simulator that models satellite motion around Earth using Newtonian gravity and an optional atmospheric drag force. The atmosphere is represented with a simple exponential density model based on altitude, which allows a direct comparison between stable orbits with drag turned off and decaying orbits with drag turned on. Over time, I track altitude, speed, and total mechanical energy to show the difference between numerical error and real energy loss caused by drag. A major part of this project was improving the code so the motion looks smooth and physically reasonable and turning the results into clear 2D and 3D visualizations. The results show that even a simplified model can reproduce realistic-looking orbital decay and reentry behavior, while also making it clear what additional physics would be needed for more accurate predictions.

Computational Pipeline for High-Quality Temporal Interference Simulation in Non-Human Primates
Presenter: Christopher De La Cruz
Faculty Sponsor: Sumientra Rampersad
School: UMass Boston
Research Area: Physics
Location: Poster Session 2, 11:30 AM - 12:15 PM: Room 163 [C24]

Temporal interference stimulation (TIS) is a neuromodulation technique that relies on the interference of high frequency electric fields to generate spatially localized modulation within brain tissue. While multiple computational modeling approaches have been successfully applied to human brain models, accurate modeling in non-human primates remains limited. This study presents the development of a computational pipeline to construct an accurate realistic non-human primate head model, specifically a Macaca mulatta head model, and be able to simulate tTIS targeting different regions of the brain model. 

Modeling non-human primates presents distinct challenges. Low-resolution MRI data, incomplete cranial coverage, and potential metal hardware are some of the main issues that we face. To address those issues, we used a high-resolution macaque atlas to construct the anatomical model, and we reconstructed the electrodes from our own images. Using automated and manual techniques, MRI images were segmented into scalp, skull, gray matter, white matter, and cerebrospinal fluid regions.. 

The completed model will then be converted into a finite element model and imported into SCIRun where we can simulate the electric field of each electrode pair and compute temporal interference fields across varying electrode current ratios. This will allow us to evaluate field strength in the target area, optimization of stimulation parameters, and spatial focus. This pipeline will establish a methodological foundation for realistic non-human primate neuromodulation modeling, and it will provide a transferable framework for future stimulation and neuroscience research. 

Simultaneous Inference of the Higgs Boson Width and Couplings Using the ATLAS Off-Shell Higgs Boson Production Measurement in the H* → ZZ Decay Channel
Presenter: Sejla Kuldija
Faculty Sponsor: Rafael Coelho Lopes de Sa
School: UMass Amherst
Research Area: Physics
Location: Poster Session 3, 1:15 PM - 2:00 PM: Campus Center Auditorium [A3]

This poster presents a study using the ATLAS measurement of the Higgs boson width published with the full Run-2 dataset at √s = 13 TeV (140 fb-1), based on the off-shell production measurement in the H → ZZ decay channel. In this study, we re-evaluate the hypotheses on the Higgs couplings used by ATLAS to measure the total Higgs boson width using the κ-framework. The measured Higgs boson width and its associated uncertainty are evaluated under different assumptions for the coupling of the Higgs boson to gluons κg and to vector bosons κV. The study combines measurements in the H* → ZZ → 4ℓ and H* → ZZ → 2ℓ2ν decay channels and reinterprets them as indirect constraints on the Higgs boson branching ratio to BSM states using the H → ZZ channel only. The results are consistent with Standard Model expectations while still allowing room for potential new-physics contributions to the total width. Full profile likelihood results are presented using the neural simulation-based inference technique.
Simulations of Tumor Treating Field Therapy in French Bulldogs
Presenter: Alexander John Vatousios
Faculty Sponsor: Sumientra Rampersad
School: UMass Boston
Research Area: Physics
Location: Poster Session 3, 1:15 PM - 2:00 PM: Campus Center Auditorium [A9]

For patients diagnosed with glioblastoma multiforme (GBM), there are limited treatment options, often restricted to surgical resection, radiation therapy, and chemotherapy. Tumor treating field therapy (TTF) has emerged as a noninvasive treatment in humans, applying electrodes to the scalp and delivering alternating electric fields to disrupt mitotic activity in cancer cells. However, the current transcranial TTF therapy faces limitations, including reduced field penetration in deep rooted tumor regions, requiring a high applied current, as well as quality of life burdens such as continuous head shavings. This study investigated the feasibility of intracranial tumor treating field therapy (iTTF) as an alternative approach for improving electric field delivery to deep-seated GBM. Given the pathological similarities between canine and human GBM we will investigate iTTF in French Bulldogs as a potential translational pathway toward human clinical  application. Using geometric models derived from MRI scans conductivity values from the literature, we ran simulations in SCIRun and MATLAb to determine electric field strengths in the tumor areas. Using the results of these simulations, we identified electrode placements that maximized field strength while minimizing applied current. Using intracranially rather than transcranially placed electrodes offers multiple benefits including increased strength and duration in which the therapy is applied, a lower current, as well as other quality of life improvements. 
Smectic Liquid Crystal Interfaces: A Toolkit for the Biosensing of Amphiphilic Molecules
Presenter: Luke Sean Riley
Faculty Sponsor: Mohamed Amine Gharbi
School: UMass Boston
Research Area: Physics
Location: Poster Session 3, 1:15 PM - 2:00 PM: Campus Center Auditorium [A57]

Liquid crystal (LC) biosensors have garnered much interest over recent years due to their high sensitivity, rapid visual changes, and ease of functionalization for detecting specific target molecules.  Amphiphilic molecules such as phospholipids and membrane proteins are important biomarkers for human health, and past studies have shown the effects of amphiphilic molecules at nematic LC interfaces, which have been promising for the possibility of LC biosensors.  However, there is a lack of research on the effects of amphiphilic molecules at smectic LC interfaces.  The smectic phase can create interesting observable defects called focal conic domains (FCDs), unlike the nematic phase, which offers the ability to create quantitative results alongside qualitative ones.  Our experiments involve us using the thermotropic LC, 8CB (4’cyano-4’octylbiphenyl) in its smectic phase and checking its feasibility as a biosensor for amphiphilic molecules.  We checked the response from different concentrations of amphiphilic molecules, such as phosphatidylcholine and cetrimonium bromide (CTAB), and we checked the feasibility of a smectic LC biosensor for each molecule, as well as the lowest detectable concentration.  Our experiments have shown a correlation between the presence of FCDs as a function of concentration and as a function of time of contact at the interface.  contribute to the understanding of molecular interactions at smectic LC interfaces and demonstrate the potential of smectic LC in biosensing applications.

Experimental Quantum Information Science with Cold Alkaline Atoms
Presenter: Camil Bernal
Faculty Sponsor: Hoang Van Do
School: UMass Boston
Research Area: Physics
Location: Poster Session 3, 1:15 PM - 2:00 PM: Campus Center Auditorium [A86]

We are in the era of noisy intermediate-scale quantum computing (NISQ), where fault‑tolerant quantum devices are not yet available due to noise and decoherence. Our aim is to create a Rydberg single‑atom array to further probe information loss during decoherence and to explore alternative pathways for preserving information in robust quantum technologies. We will present our latest experimental progress on achieving a cold magneto‑optical trap (MOT).
Numerical Simulation For Modeling and Mitigating Parasitic Conduction in NbTi–Cu Leads
Presenter: Connor D. McLaughlin
Faculty Sponsor: Scott Hertel
School: UMass Amherst
Research Area: Physics
Location: Poster Session 4, 2:15 PM - 3:00 PM: Campus Center Auditorium [A58]

In ultra-low temperature cryogenics, such as dilution refrigeration, the primary challenge for power delivery is managing heat loss. While superconducting wires offer a zero-resistance approach for delivering power, they also act as parasitic thermal links. This project develops a C++ simulation to quantify and minimize these parasitic thermal links in a composite wire consisting of Niobium-Titanium (NbTi) and Copper (Cu). 

Because material properties at cryogenic scales vary by orders of magnitude, an implicit Crank-Nicolson method is used to solve the heat equation. This accounts for the temperature-dependent evolution of material properties across discrete nodes. To address the numerical challenges of numerical stiffness caused by sharp material interfaces (where NbTi meets the Copper), Jacobian-based smoothing functions are implemented. This ensures the simulation converges on a stable steady-state solution without typical numerical oscillations.

The final testing suite allows for the visualization of power-removal requirements at each refrigeration stage. By iterating on the placement of heat anchors and the volumetric ratio of NbTi to Copper, this tool provides a path to optimize lead assemblies and reduce the cooling power required to maintain base temperatures in complex cryogenic systems.


Engineering the Motion of Self-Propelled Particles on Surfaces
Presenter: Ali Imraan
Faculty Sponsor: Manasa Kandula
School: UMass Amherst
Research Area: Physics
Location: Poster Session 4, 2:15 PM - 3:00 PM: Campus Center Auditorium [A60]

Chemically active Janus colloids - microscale particles with two distinct faces, one catalytic platinum and one inert - offer a controllable model for active fluids. Unlike Brownian diffusion, active particles exhibit out-of-equilibrium life-like motion. When placed in aqueous suspensions of hydrogen peroxide, asymmetric surface reactions generate ionic gradients that create a local electric field and drive electroosmotic slip within the electrical double layer, leading to electrokinetic self-propulsion. In this regime, the particle’s zeta potential sets the slip polarity and therefore orientation: negatively charged swimmers translate with the inert face leading, whereas a positive surface charge reverses propulsion and yields cap-first motion. This study investigates how catalytic cap size and light-controlled surface charge influence propulsion strength and orientation dynamics. We hypothesize that increasing cap area increases reaction flux and thereby strengthens propulsion, while also altering orientation dynamics through changes in near-surface flows. Further, we postulate that illumination shifts the particle’s zeta potential and therefore the magnitude and polarity of electrokinetic slip, allowing for real-time control of these active colloids. To this end, we perform microscopy experiments to investigate the motion and use in-house developed codes for quantitative analysis of the dynamics. Our results show thickness dependent shifts in speed distributions and cap-velocity alignment, and illumination allowing for modulation of motility. These findings establish experimentally tractable control of active transport through coupled tuning of catalytic geometry and surface charge, providing a framework for programmable micromotor design.
Buoyancy Compensation Using Magnetic Levitation of Inertial Particles in Fluid Flows
Presenter: Milo M. Van Mooy
Faculty Sponsor: Varghese Mathai
School: UMass Amherst
Research Area: Physics
Location: Poster Session 4, 2:15 PM - 3:00 PM: Campus Center Auditorium [A62]

Gravity-driven flows such as surface waves, boundary-layer currents, and stratified motions govern the transport of finite-size particles in many natural and engineered environments. An increasingly relevant example is the transport of microplastic pollution in the ocean. Despite their small size, these particles have non-negligible mass which in part determines their behaviour within the waves and currents of the ocean. In order to assess health risks and find plastic removal solutions we must understand and predict where these particle aggregate and why. In general, particle trajectories reflect both gravity-related effects (buoyancy and settling) and inertia-related effects (finite response time and drag), which are coupled. This means that in practice, field observations often cannot disentangle these distinct effects. Here we develop a magnetic levitation (maglev) approach that applies an approximately constant vertical magnetic force to designed spherical particles in a water flow. The novel magnetic levitation system uses the superposition of precisely engineered coaxial permanent magnet disks, creating a  roughly 10 cubic centimeter volume within which the net buoyancy force on particles is compensated to within 10%, while horizontal forces remain below 10% of the vertical compensation. We measure the field and discuss practical limitations including tolerances and time-dependent demagnetization. This framework provides a reproducible route to buoyancy-cancellation experiments in gravity-driven flows with relevance not just to microplastics, but to numerous problems in environmental transport.

RELATED ABSTRACTS

Long-Term Characterization of Prototype Silicon Photomultipliers in Liquid Xenon
Presenter: Zihan Rao
Faculty Sponsor: Andrea Pocar
School: UMass Amherst
Research Area: Physics
Location: Poster Session 6, 4:15 PM - 5:00 PM: Campus Center Auditorium [A40]

The search for neutrinoless double beta decay is one of today’s most compelling challenges in physics, as its observation would confirm the Majorana nature of neutrinos and provide crucial insights into their absolute masses. A fermion is said to be a Majorana when it coincides with its antiparticle. If this were true for neutrinos, the discovery of neutrinoless double beta decay could help to reveal the reason why we live in a matter-dominated universe in which antimatter is almost absent. In support of this quest, detectors such as the nEXO (next Enriched Xenon Observatory) experiment are under development. My research focuses on the operation and analysis of a Liquid Xenon Cryogenic System at Andrea Pocar’s lab at UMass Amherst - a xenon liquefaction prototype setup run as part of the nEXO R&D. A critical function of the system is to test silicon photomultipliers (SiPMs) in liquid xenon, thereby supporting the development of detectors for nEXO.

My contributions span the operation of the LXe cryogenic system and the waveform analysis of silicon photomultipliers. An SiPM is a highly sensitive light detector capable of resolving single photons. Key performance parameters include single-photon (SPE) resolution, breakdown voltage, dark count rate, and correlated avalanche probability. 

In my work, I characterize the long-term behavior of prototype SiPMs by measuring breakdown voltage and gain in vacuum, gaseous nitrogen, and most importantly in liquid xenon (LXe), and study how xenon scintillation light propagates in an nEXO-like detector configuration.

RELATED ABSTRACTS

Study of Entropic Surface Anchoring for Lyotropic Liquid Crystals
Presenter: Juliet Zhu
Faculty Sponsor: Shuang Zhou
School: UMass Amherst
Research Area: Physics
Location: Poster Session 6, 4:15 PM - 5:00 PM: Room 163 [C12]

Lyotropic liquid crystals (LCs) are formed by suspensions of a wide range of anisometric particles, such as nanorods, colloidal disks, and macromolecular assemblies. At high concentration, the particles align with each other to reduce mutual exclusion volume and gain more translational entropy, resulting in a nematic orientational order. Controlling surface alignment of lyotropic LCs is critical in studying their physical properties and engineering them in applications, and is therefore a key next step of the research. The purpose of this study is to understand the surface anchoring energy of lyotropic liquid crystal due to nanopillar arrays on the substrate. We apply a magnetic field to distort the director field in the cell and observe the Frederiks transition. Applying a horizontal magnetic field in a vertically aligned cell will induce a competing factor of the free energy, where above a critical field, the director prefers to align with the magnetic field in the bulk. By measuring the threshold field as a function of cell thickness, we extract the extrapolation length associated with finite anchoring and determine quantitatively the anchoring energy using the bend elastic constant of the system. Our results will establish frameworks and techniques for characterizing surface anchoring in lyotropic liquid crystals and demonstrate the effectiveness of nanopatterned substrates in producing controllable homeotropic alignment.
Motion of Crithidia fasciculata as Driven by Singular Forward-Moving Flagellum
Presenter: Lauren Helena Daniel Moran
Faculty Sponsor: Shuang Zhou
School: UMass Amherst
Research Area: Physics
Location: Poster Session 6, 4:15 PM - 5:00 PM: Room 163 [C13]

The dynamics produced by the swimming of parasitic Crithidia fasciculata are understudied, making these microorganisms crucial for understanding the underlying physics of their locomotion. C. fasciculata are known as “puller” swimmers, meaning their front-mounted flagellum lead their motion, distinct from better known “pusher” swimmers such as E. coli and spermatozoa with the propeller behind the body. Through defocused particle imaging microscopy tracking of dyed fluorescent dyed particles attached to different parts of the body and flagellum of C. fasciculata, we obtain three dimensional data to analyze their trajectory and body dynamics. Particle image velocimetry (PIV) are used to better understand resulting flow fields surrounding the swimmers. This study of C. fasciculata paves the way for more knowledge about parasites and their motility, flagellar structure, and locomotion in a diverse array of circumstances. Furthermore, understanding the locomotion of C. fasciculata can support vector analysis illustrating parasite transmission dynamics, since our findings can lead to a clearer view of their colonization and transmission, as in other pathogenic species.