Abstracts — MICRO

MICRO Abstracts

Would you like to learn more about the sort of research MICRO students work on? The abstracts below describe the work MICRO students have performed recently. You can find them below organized by research topic.

Biomaterials

Image by Julia Higuchi, Jan Mizeracki CC-BY 4.0

Rachel Myers

Techniques for high-throughput validation of the genomic drivers of nanoparticle interactions in cancer cells

Rachel Myers^1,2, Dr. Joelle P. Straehla^2,3, Dr. Paula T. Hammond^2,3,4

¹ Department of Chemical, Biochemical and Environmental Engineering, University of Maryland Baltimore County; Baltimore, MD
² Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology; Cambridge, MA
³ Broad Institute of MIT and Harvard; Cambridge, MA, USA
⁴Department of Chemical Engineering, Massachusetts Institute of Technology; Cambridge, MA

Cancer nanomedicine is a promising venture for the delivery of therapeutics to patients with cancer, as nanoparticles (NPs) can be designed to increase tissue specificity and efficacy for a range of therapeutic cargos. Pooled nanoPRISM screening has identified numerous candidate genes that can regulate NP-cancer-cell interactions, but high-throughput approaches are needed to evaluate them simultaneously1. Based on this established work that nominated candidate genes, we sought to validate their role in NP delivery by using a CRISPR/Cas9 pooled screen. In this screen, each cell has a single gene de-activated, and next generation sequencing can be used to determine the relative enrichment of genes in populations defined by their extent of NP association. Here, we discuss screen design and data analysis strategies to determine which genes were associated with enriched or depleted NP interactions after performing a pooled knockout screen on BT245 human glioma cells treated with layer-by-layer nanoparticles. Layer-by-layer nanoparticles allow for the tunable assembly of nanoparticles and allow us to decouple the influences of the NP core or the layered surface chemistry on NP-cell interactions. Here, we dosed the cells with two NP formulations: 1) a bare liposomal NP, and 2) a liposomal NP layered with hylauronic acid. The hylauronic acid layer was of interest because cancer cells commonly express hylauronic acid receptors, also known as CD44 receptors. We found that several genes of varying functionalities were significantly involved in NP uptake upon knockout in the cells. This genomics-based approach to study NP-cell interactions provides an efficient means to interrogate biologic regulators of NP delivery to cancer cells.

Reference: ¹Boehnke N, Straehla JP, Safford HC, Kocak M, Rees MG, Ronan M, et al. Massively parallel pooled screening reveals genomic determinants of nanoparticle delivery. Science 2022;377:eabm5551. https://doi.org/10.1126/science.abm5551.

Nicholas Layman

Using Dynamic Time Warping for Identification of Peptides in Affinity Selection Mass Spectrometry

Nicholas Layman, Somesh Mohapatra, Rafael Gómez-Bombarelli

One of the early stages of drug discovery is the identification of a protein's specific binders to aid in localizing medicine absorption and to ensure effective treatment. Identifying specific binders one at a time is straightforward but extremely time-consuming so we use affinity selection - mass spectrometry (AS-MS) to test many potential binders at once and obtain a collection of suitable binders. To make this process more robust we repeat it a few times per protein and combine these trials into one average dataset. These datasets cannot be combined naively due to small temporal variations in a filtering stage of AS-MS. To overcome these challenges, I implemented and optimized dynamic time warping methods to obtain and analyze a typical dataset. The fully functional and accessible code base I developed enables process automatization as the first step towards systematic binder selection from any AS-MS dataset.

Melanie Andrade-Muñoz

Investigation of the Hierarchical Progression of Enamel Densification

Melanie Andrade-Muñoz^1,2, Ethan Suwandi², Derk Joester²

¹ Department of Neuroscience and Psychology, University of North Carolina at Chapel Hill, Chapel Hill, NC
² Department of Materials Science and Engineering, Northwestern University, Evanston, IL

Enamel is the most highly mineralized tissue in the human body, composed primarily of a complex hierarchical structure of highly elongated hydroxyapatite crystallites. Enamel formation, or amelogenesis, occurs in two main stages: the secretory stage, where the core structure of crystallites is laid down, and the maturation stage, where the crystallites densify as the organic protein matrix is degraded. This unique process distinguishes enamel formation from typical crystal growth methods, involving the interplay between crystallite growth and matrix degradation. Despite its critical role in enamel strength and durability, no existing microstructure model fully captures the hierarchical progression of enamel densification during amelogenesis. This gap limits our ability to understand and address disruptions in enamel formation, such as those seen in congenital diseases like Amelogenesis Imperfecta (AI).

To bridge this gap, we develop a computational model using FiPy, an object-oriented partial differential equation (PDE) solver, to simulate crystallite growth via phase-field methods. FiPy enables a microstructure-level simulation of anisotropic crystallite-matrix interactions, linking microstructure evolution to experimental density data (e.g., microCT) and providing mechanistic insights into hypomineralization disorders like AI. Incorporating anisotropic surface energy and unique starting structures and assuming a constant boundary growth rate, we can replicate key features like crystallite elongation and packing. We calculate porosity, major/minor axes, and crystallite orientation, enabling connections between microstructure (crystallites) and macrostructure (density). This framework lays the groundwork for understanding hypomineralization disorders, such as AI, and provides a foundation for exploring the effects of biological variables on enamel formation. Ultimately, this research advances our understanding of enamel development and its disruption in disease states, offering potential pathways for diagnosing and intervening in enamel-related disorders.

Materials for Energy Applications

Image by Tennen-Gas CC BY-SA 3.0

Jon-Edward Stokes

Jon-Edward Stokes¹, Richard Church², Professor A. John Hart³
¹Department of Physics and Astronomy, Howard University, Washington, DC
²Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA
³Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA

Finite element analysis of lithium-ion batteries with optimized energy and power density via a 3D honeycomb architecture

Despite significant advances the existing planar lithium-ion battery configuration is approaching its theoretical energy density limit. One strategy to obtain additional increases in energy density is to play with the cell geometry to minimize the mass and volume contributions of inactive materials. Typically, this minimization is achieved by increasing electrode thickness. However, this increased thickness results in increased Li-ion diffusion distances, thereby decreasing the power density. For this reason, planar cells are viewed as having an inherent trade-off between energy and power density. This trade-off may be overcome by further modifying the full cell geometry to incorporate 3D, such as interpenetrating or interdigitated, architectures. 3D architectures enable both high energy and power density due to a higher surface area for the same volume and footprint of cell. As a result, a given volume of active material can be distributed with a lower electrode feature thickness which decrease the Li-ion diffusion distance compared to an equivalent planar cell. In this study, we use perform finite element analysis simulations using COMSOL to predict the performance of 3D batteries with a unique 3D architecture built upon honeycomb-patterned vertically aligned carbon nanotubes (CNTs). Here, CNTs have been chosen for their excellent mechanical strength and electrical conductivity, allowing them to serve as a 3D scaffold for cell fabrication. These discharge simulations will determine the properties required to make these 3D CNT-based cells competitive with existing lithium-ion designs in terms of CNT spacing and height, and will be incorporated into the experimental work trying to produce 3D CNT-based full cells.

Temiloluwa Akande

Advisors: Rafael Gomez-Bombarelli and Pablo Leon

The effects of different functionalization of standard electrolyte solvents for lithium batteries using computational workflow and molecular dynamics.

A lithium-ion battery is a rechargeable battery that uses the reduction of lithium ions to store, generate and transfer energy. The lithium-ion battery set up consists of the anode typically carbon based such as graphite. The cathode, the positive electrode, is usually a metal oxide. The electrolyte is the electrolyte would consist of the lithium salt and organic solvent. Typical electrolytes used are highly volatile and impractical for many applications. Although a wide array of electrolytes exists, carbonate-based electrolytes have been used in commercial Li-ion batteries for three decades and are a natural and practical choice to replace common type. Common types of electrolytes are carbonate-based electrolyte with cyclic carbonates (e.g., ethylene carbonate (EC), and propylene carbonate (PC)), linear carbonates (e.g., ethyl methyl carbonate (EMC), diethyl carbonate (DEC), and dimethyl carbonate (DMC)).

The aim of this project is to use molecular dynamics to understand the effect of altering the concentration of ions and solvents of the electrolyte of lithium-ion batteries. To do this, an array of boxes each comprising of a varying number of solvents (eg ethylene carbonate (EC)) and a varying number of ions (lithium or hexafluorophosphorus PF6) would be made. The number of ions and solvents for each box created would also vary by a factor of 10 and 2 to consider what happens to boxes made entirely too small. After the boxes are created, they would be equilibrated through a process where the different solvent and ion molecules eventually condensing into a liquid-phase box from a gas-phase box.

The next phase of the project would consider how the boxes created are affected by Temperature, speed, density all as a function of time. The results from such analysis would give insight into designing the boxes for molecular simulation which would in turn give insight into creating carbonate electrolytes for lithium ion batteries.

Tobi Majekodunmi

Finite Element Analysis of a Three-Dimensional Solid-State Battery

Tobi Majekodunmi¹*, Richard Church², Professor A. John Hart³

¹Department of Mechanical Engineering, University of Maryland, Baltimore County, Baltimore, MD
²Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA
³Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA

Lithium-ion batteries are ubiquitous within today’s energy storage applications, particularly in the emerging field of electric vehicles. To be competitive with existing transportation technologies and overcome the “range anxiety” felt by many consumers, electric vehicles require batteries with higher energy densities than what is currently available. Without exploring other chemistries, the simple solution is to utilize thick electrodes. However, increasing electrode thickness in planar cell designs comes at the expense of power density, which negatively impacts the car’s acceleration. MIT’s Mechanosynthesis Group is developing a solid-state battery for Lamborghini’s pilot electric supercar. This battery utilizes a 3D, interdigitated geometry and highly conductive carbon nanotubes to simultaneously increase energy density and power output. This project performed discharge simulations with COMSOL modeling software to guide experimental design and fabrication by analyzing how the battery’s geometry, ionic conductivity, and electronic conductivity impact performance.

Griheydi Garcia

A High-throughput Study of Apatite Minerals For Energy Applications

Lead apatites have received a lot of attention due to the recent report of a room temperature superconductor, LK-99, Pb₉Cu(PO₄)₆O, and have been in the spotlight ever since. LK-99 represents just one facet of a much broader category of minerals collectively known as apatites. The Apatite supergroup encompasses diverse subgroups, each characterized by distinct chemical compositions, with a general structural formula M12M23(TO4)3X, where M is any 2+ metal, T is Si⁴⁺, V⁵⁺, P⁵⁺, As⁵⁺, etc. and X is OH^-, F^-, Cl^-, Br^-, I^-, and O^2-. The apatite family encompasses a wide range of compositions and shows various characteristics that could be useful for potential energy materials. For the first time, we’re showing that this class of minerals could be potential energy materials, and showcase a wide range of bandgaps ranging from metals to semiconductors to insulators and they possess fairly simple structures that are known to be synthesizable in the case of lead apatites. We tend to identify new stable members of this family through our high-throughput search and find their potential applications. In this study, we unveil the potential of apatite minerals as energy materials, emphasizing the diverse bandgap range they offer, ranging from metals to semiconductors to insulators. Through a comprehensive high-throughput search, we aim to identify new thermodynamically stable members within this mineral family and explore their potential applications. Leveraging the prototype structure of lead apatites with charge-similar elements (Z=2⁺), our research reveals a spectrum of compounds encompassing insulators, semiconductors, and metals. These findings open avenues for potential applications in energy-related fields, including photovoltaics and thermoelectrics.

Photonic & Magnetic Materials

Image by Quasic CC BY-SA 2.0

Anastacia De Gorostiza

Anastacia De Gorostiza¹, Katherine Stoll², Henry Carter³, Yudong Yang³, Jonathan Sessler³, Zachariah Page³, Samuel Serna-Otálvaro^2,4, Anuradha Agarwal²
¹ McKetta Department of Chemical Engineering and Texas Materials Institute, The University of Texas at Austin, Austin, TX 78712-1062, USA
² Department of Materials Science and Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA
³ Department of Chemistry, The University of Texas at Austin, Austin, Texas 78712-1224, United States
⁴ Bridgewater State University, Physics Department, 131 Summer St., Bridgewater, MA 02324, USA

Sensitive and Selective On-Chip Methane Detection

Rising methane emissions due to human industrial activity have increased interest in understanding and preventing such emissions as they pertain to climate change. However, one prevailing challenge in quantifying methane emissions from industrial sources is the development of a device that is both inexpensive and sensitive to concentrations below 0.1 ppm. A potential low-cost and high-efficiency material platform for methane sensing is the use of on-chip photonic sensors incorporated with a methane-sensitive polymer cladding. In this project, we present a sensitive on-chip methane sensor with a polymer cladding layer composed of a blend of styrene-acrylonitrile block copolymer and a methane-selective molecule cryptophane A. The performance of our polymer-cladded chip was compared to a control chip. Furthermore, the selectivity of cryptophane A to methane and other gases was explored.

Dawn Ford

Dawn Ford¹, Mads Weile^2,3, Frances M. Ross³, Julian Klein³
¹Department of Physics, University of Virginia, Charlottesville, VA
²Department of Physics, Technical University of Denmark, Lyngby, Denmark
³Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA

Efficient detection of defects and phases in electron microscopy images of CrSBr using deep learning

Two-dimensional (2D) magnets are an interesting class of materials in which magnetic order and atomic structure are intrinsically correlated. Controlling the atomic structure of 2D magnets provides an avenue to control spin-related phenomena with exciting new applications for quantum phase engineering. A particularly exciting material for structural modifications is the air-stable 2D magnetic semiconductor CrSBr that has weak antiferromagnetic interlayer coupling and ferromagnetic ordering in the monolayer limit, with the magnetic moments situated at the Cr atoms. Imaging CrSBr in the scanning transmission electron microscope (STEM) reveals a structural phase transformation where Cr atoms become mobile moving into interstitial sites in the van der Waals gap. The mechanism is complex and further atomistic analysis is required for better understanding of the kinetic processes that drive the transformation. A machine learning workflow can be used to effectively track and study structural changes of CrSBr when imaged under the electron beam. We use deep convolutional neural networks (DCNNs) trained on CrSBr to effectively detect atomic column positions. From these positions we determine local and global strain to quantitatively explore changes in the crystal lattice as the structural transformation evolves. From these atomistic insights we expect to learn more about the kinetics of the phase transformation, which is of great value for inspiring future electron beam guided structural modifications.

Jordan Coney

Advisor: Dr. Juejun Hu/Tushar Sanjay Karnik

Light Loss in Bending Waveguides

I will work for MIT’s MICRO program with Dr. Juejun Hu’s lab under a graduate student named Tushar Sanjay Karnik, where he will minimize the light loss from curved waveguides using the simulation tool Lumerical. The goal is to make integrated photonic circuits more miniature in size. We will look at the different techniques and parameters of carrying this out. A waveguide is a structure that guides waves. For us, they guide electromagnetic waves (think microscopic optical fiber). The electromagnetic waves propagate through the waveguide by a process called total internal reflection. When you bend a waveguide, radiation becomes lost from before the bend to after the curve. In this project, the implemented radiation sources are mid-infrared. The materials that are being used in this project are considered III-V semiconductors. The materials used are essential in integrated photonics because they can be used to make on-chip lasers. We are using an InP cladding and an InGaAs core for the waveguide. I plan on encountering challenges with scheduling as well as software issues. I have had cases where I have difficulty creating the code to carry out our simulation. Last semester I introduced the air trench to our configuration simulation, so I will not continue to expand on the depth of this and explore its physical limitations.

Joshua Chaj Ulloa

Deep Neural Networks Hyperparameters Optimization within Photonic Meta Surface Devices Simulations

Joshua Chaj Ulloa, Sensong An, JueJun Hu

Photonic systems and devices are applicable to various biomedical research applications, with the overarching scope of the field being the study of light and its applications to biosensing, metamaterial optimization, and optical engineering. However, for photonic devices to be applicable to their specific field applications, customization and specification with regards to their device geometry and optical response must be studied and troubleshot for their certain optical properties to be tailor-made. Therefore, researchers have applied the field of neural network machine learning algorithms toward the specific development and optimization of various physical, spectral, and performance characteristics of photonic devices to enhance these properties. Our project focuses on the development of neural networks that will allow us to recognize hidden patterns and correlations for the photonic meta surface devices experimental datasets. Where an overall exploration will be conducted of the effects specific hyperparameter tuning will have on the neural network functionality and performance as this tuning allows for the specific optimization of the hidden layer and overall structure of the deep neural network. The analysis will apply specific hyperparameter tuning methods such as Grid Search, Random Search, and Bayesian Optimization toward monitoring the performance of the complex neural networks developed. This unique approach provides promise toward the enhanced simulation and design development process of photonic meta surface devices toward their specific biomedical research applications.

Neal Haldar

Neal Haldar¹, Drew Weninger², Luigi Ranno², Ajay Gupta², Anuradha Agarwal²
¹College of Chemistry, University of California, Berkeley, Berkeley, CA, 94720, USA
²Department of Materials Science and Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA

Life Cycle Analysis of Photonic Chips

Increasing computational requirements in large data centers and the like has outstripped traditional electronic circuits and has required the rapid development of alternatives. At the same time, the highly polluted supply chains, in terms of carbon emissions and others, for the creation of electronic microchips has created a demand for cleaner alternatives. A potential option is an electronic photonic coupled package that minimizes the resistive heating. In this project, we are performing a life-cycle analysis of one such photonic chip and evaluating its environmental impact in comparison with a standard silicon microchip. The resulting data will be published as a reference to evaluate future designs in this field and provide an acceptable benchmark.

Axel Magaña Ponce

Axel Magaña Ponce¹, Abhiram Devata², Prof. David Barton²
¹Department of Physics and Engineering Physics, Elmhurst University; Elmhurst, IL
² Department of Material Science and Engineering, Northwestern University; Evanston, IL

Lithium Niobate Ring Resonators for Matrix Elements

As the demand for artificial intelligence (AI) continues to rise, its energy consumption challenges future innovation and deployment. In recent years, photonic computing has emerged as a solution that utilizes light instead of electrons to process data with greater speed and efficiency. To further augment these benefits with the neuromorphic algorithms necessary to train AI systems, nonvolatile matrix elements must be represented on a photonic platform. Due to its desirable optical properties, we aim to investigate these elements on thin-film lithium niobate (TFLN), a candidate material for scalable integrated photonics. Ring-resonator-based elements will be designed for their ability to filter wavelengths, store energy, and enable compact, low-loss devices. These elements will be optimized using Lumerical, a computational software, to simulate how ring resonator geometries impact critical performance metrics such as free spectral range and quality factor. By exploring a comprehensive parameter space, we will identify optimal designs for robust operation across multiple wavelengths while addressing challenges related to scalability and fabrication.

Polymers & Soft Matter

Image by Minutemen CC BY-SA 3.0

Gabrielle Wood

Wood, Gabrielle N¹, Chazot, Cécile AC²
¹Department of Chemical Engineering, Lewis K. Downing College of Engineering and Architecture, Howard University, Washington, DC.
²Department of Materials Science and Engineering, Robert R. McCormick School of Engineering and Applied Science, Northwestern University, Evanston, IL.

Solubility Parameter Predictions for Textile Recycling via Selective Dissolution

The solubility parameter is a numerical value that captures the degree of interaction between materials from their individual cohesive energy density. The solubility parameters of different species can be compared to assess their miscibility under the principle that “like dissolves like”. In particular, group contribution theory enables numerical prediction of solubility parameters assuming that the properties of organic molecules, such as monomers, polymers and solvents, can be obtained by adding the contribution of their individual constituents and functional groups. In this project, a MATLAB program was created to calculate solubility parameters from two frameworks of group contribution theory, the first presented by Hansen, the other, expanding on Hansen, by Stefanis and Panayioutou. We then implemented this theory to the design of polymer processing and recycling strategies, specifically in the case of currently non-recyclable elastane-containing textiles. Based on solubility parameter calculations, we investigated multiple strategies for the processing and recycling of elastane filaments to improve separability and isolation through selective dissolution. We explore how solubility and non-covalent interactions can be leveraged to facilitate recycling without sacrificing interfacial adhesion with blended fibers. This work has the potential to develop alternative blended elastane fibers and identify scalable chemical recycling methods based on an environmentally-friendly selective dissolution method.

Nolan James Murphy-Genao

Nolan James Murphy-Genao¹, Carl Thrasher², Robert MacFarlane³

¹Department of Chemical & Biomolecular Engineering, University of Connecticut, Storrs, CT 06269, USA

^2,3Department of Materials Science and Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA

Self-assembly of nanostructures is an innovative methodology where atoms, nanoscale building blocks, or molecules organize themselves in response to stimuli (e.g., thermal, light) into complex ordered structures or patterns with nanometer features without any notable human intervention. This facile mechanism offers surfeit potential as it is currently a promising practical, economical, and proliferative avenue for nanofabrication. Unfortunately, modern self- assembly technologies do not possess spatial control over rudimentary or intricate architectures. This gratuitous obstacle is a significant detriment as successfully obtaining complete regulation of polycrystalline structures could be an avant-garde discovery. However, forming superlattices comprised of DNA nanoparticles that serve as programmable atom equivalents (PAEs) that perform on substrates offers new possibilities for manipulating hierarchical structures by guiding their crystallographic orientation and placement. This work uses substrates patterned via optical lithography to explore the nucleation and growth behavior of PAE crystals on substrates, control the shape of polycrystalline assemblies, and the location and orientation of large single crystals with the addition of computer-guided characterization of external properties to analyze the progress.

Zorah Williams

Zorah C. Williams¹, Dr. Cécile Chazot², Simona Fine³

¹Department of Chemical, Biochemical and Environmental Engineering, University of Maryland Baltimore County, Baltimore, MD

^2,3Department of Materials Science and Engineering, Northwestern University, Evanston, IL

Predicting Phase Behavior in Electrochromic Polymer Blends Using Flory-Huggins Theory

Electrochromic materials, which change color when voltage is applied, have promising applications in sensors and electronic materials. Traditional electrochromic systems rely on dyes that change color, but these dyes have drawbacks, including environmental instability and pollution from wastewater contamination. An alternative approach involves using structurally-colored polymer composites, which create color through nanoscale molecular arrangements rather than dyes. These materials offer greater stability, sustainability, and the ability to detect multiple stimuli.

This study explores the integration of poly(3,4-ethylenedioxythiophene) (PEDOT), a conductive polymer often used in electrochromic materials, with cellulose-based liquid crystals, specifically ethyl cellulose (EC) and poly(acrylic acid) (PAA). These cellulose-based materials can self-assemble into ordered liquid crystalline structures (mesophases) that produce structural color. By incorporating PEDOT into these EC and PAA mesophases, we aim to create electrochromic materials that are both highly conductive and visually tunable.

We use a MATLAB-based computational model to calculate the free energy of mixing for binary polymer systems to predict how these polymers mix and form a stable blend. Specifically, we analyze EC-PEDOT and PEDOT-PAA blends using Hansen solubility parameters, group contribution theory, and Flory-Huggins theory—methods that help determine how well different polymers mix at the molecular level. These calculations allow us to generate phase diagrams, which indicate the conditions, such as molecular weight and temperature, that promote uniform mixing and prevent phase separation. This study provides guidance for experimental synthesis by identifying optimal compositions, ensuring PEDOT is effectively incorporated for enhanced electrochromic behavior.

This study contributes to the rational design of electrochromic polymer blends by developing a computational framework to predict phase behavior. Future work will expand this approach to other cellulose derivatives and conductive polymer systems, further supporting the creation of sustainable, multifunctional materials for environmental sensing and other advanced applications.

Metallurgy

Eyobel Haile

Metal Additive Manufacturing by selective laser melting (SLM) is a promising technology for large scale manufacturing, but is also a time consuming, expensive, and labor intensive process. From this perspective, it is important to make sure that the printing process goes as smoothly as possible and produces consistent parts to avoid additional processing steps or failed prints. In this project, in-situ quality control techniques were implemented in order to inspect the structural integrity of printed parts during the process. The hardness data of certain materials of interest was collected and used for determining the local quality of printed parts while controlling for all other variables. Linking measured hardness to temperature gradient, porosity and part defects enable identification of localized defects in the print. Such a method opens new opportunities for reducing the occurrence of failed prints and improve printing quality.

Tobi Majekodunmi

Finite Element Analysis of a Three-Dimensional Solid-State Battery

Tobi Majekodunmi¹*, Richard Church², Professor A. John Hart³

¹Department of Mechanical Engineering, University of Maryland, Baltimore County, Baltimore, MD
²Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA
³Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA

Lithium-ion batteries are ubiquitous within today’s energy storage applications, particularly in the emerging field of electric vehicles. To be competitive with existing transportation technologies and overcome the “range anxiety” felt by many consumers, electric vehicles require batteries with higher energy densities than what is currently available. Without exploring other chemistries, the simple solution is to utilize thick electrodes. However, increasing electrode thickness in planar cell designs comes at the expense of power density, which negatively impacts the car’s acceleration. MIT’s Mechanosynthesis Group is developing a solid-state battery for Lamborghini’s pilot electric supercar. This battery utilizes a 3D, interdigitated geometry and highly conductive carbon nanotubes to simultaneously increase energy density and power output. This project performed discharge simulations with COMSOL modeling software to guide experimental design and fabrication by analyzing how the battery’s geometry, ionic conductivity, and electronic conductivity impact performance.

Pablo Luna Falcon

Machine Learning Prediction of the Al-Ce-Ni-Mg System Mechanical Properties

Pablo Luna Falcon¹, Jie Qi², David Dunand³

¹Department of Mechanical Engineering, University of Arizona, Tucson, AZ
²Department of Material Science and Engineering, Northwestern University, Evanston, IL

Aluminum cast alloys based on the eutectic Al-Ce-Ni-Mg system are a relatively new system. The formation of continuous Al₁₁Ce₃ and Al₃Ni phases, coupled with the incorporation of Mg into the Al matrix, results in enhanced alloy strength and improved creep resistance through precipitate strengthening and solid solution strengthening mechanisms. Additionally, the low solubility and diffusivity of Ce in Al contribute to exceptional resistance to high-temperature coarsening. The eutectic structure of these alloys also ensures excellent castability. Although the promising properties for this alloy system has been demonstrated for certain compositions, the traditional trail-and-error research approach is time-consuming and expensive. Recognizing the inherent complexity of alloy development, this study proposes a novel machine learning (ML) model employing a gaussian process regression algorithm for yield strength and hardness prediction within the Al-Ce-Ni-Mg system. The ML models use composition and alloy processing methods as features, demonstrating low prediction errors of 69 and 18 MPa for hardness and yield stress. From the scientific perspective, strengthening effects from individual elements and processing methods are studied. Ongoing efforts involve refining the model performance with plans to incorporate a larger alloy system. Predictions will be experimentally validated, and undiscovered high-performance alloys will be designed by the ML models.

Simone Lang

High entropy alloys for corrosion resistance applications

Simone Lang ¹, Yifan Cao², Rodrigo Freitas²

¹Department of Chemistry and Biochemistry, Texas Woman’s University, Denton, Texas 76204, USA
²Department of Material Science and Engineering, Massachusetts Institute of Technology,Cambridge, Massachusetts 02139, USA

High-entropy alloys (HEAs) are formed by mixing three or more elements in nearly equal proportions, resulting in significant chemical complexity. This complexity makes it challenging to experimentally characterize chemical short-range order (SRO) and its effects on material properties [1]. To overcome such difficulty, computational methods have been employed to characterize SRO in complex alloys over the past few years [2,3]. Predictive simulations, for example, can accurately model the evolution of surface chemistry in HEAs, aiding their application in catalysis and corrosion-resistant structures [4]. CrCoNi has been suggested as a promising candidate for corrosion resistance due to its high chromium content. In its bulk phase, previous studies have suggested that the Cr-Cr pairings are highly disfavored, causing chromium segregation on the 2nd nearest neighbor distances and potential loss of corrosion resistance [4]. However, the impact of SRO on CrCoNi’s surface corrosion resistance remains unclear. This study investigates the catalytic surface interactions of CrCoNi by simulating and characterizing its chemical profile near free surfaces. Hybrid Molecular Dynamics Monte Carlo (MDMC) simulations were used to analyze SRO across various temperatures and concentration profiles throughout the material. Warren-Cowley parameters were calculated at different depths of the free surface and bulk system to quantify SRO variations. Structural changes over time were visualized using the Open Visualization Tool (OVITO). Additionally, free surface energy with and without SRO was determined based on the bulk system energy and surface area to assess surface-related properties of CrCoNi [5].

References
[1] Joress, H., Ravel, B., Anber, E., Hollenbach, J., Sur, D., Hattrick-Simpers, J., Taheri, M., and DeCost, B. Why is EXAFS for complex concentrated alloys so hard? Challenges and opportunities for measuring ordering with X-ray absorption spectroscopy. Matter. (2023). https://doi.org/10.1016/j.matt.2023.09.010.
[2] Zhang, R.; Zhao, S.; Ding, J.; Chong, Y .; Jia, T.; Ophus, C.; Asta, M.; Ritchie, R. O.; Minor, A. M. Short-Range Order and Its Impact on the CrCoNi Medium-Entropy Alloy. Nature 2020, 581 (7808), 283–287. https://doi.org/10.1038/s41586-020-2275-z.
[3] Chen, W.; Hilhorst, A.; Bokas, G.; Gorsse, S.; Jacques, P. J.; Hautier, G. A Map of Single-Phase High-Entropy Alloys. Nat Commun 2023, 14 (1), 2856. https://doi.org/10.1038/s41467-023-38423-7.
[4] Cao, Y .; Sheriff, K.; Freitas, R. Capturing Short-Range Order in High-Entropy Alloys with Machine Learning Potentials. arXiv January 12, 2024. http://arxiv.org/abs/2401.06622 (accessed 2024-06-13).
[5] Bodlos, R., et al. "Energies and Structures of Cu/Nb and Cu/W Interfaces from Density Functional Theory and Semi-Empirical Calculations." Materialia, vol. 21, 2022, p. 101362, https://doi.org/10.1016/j.mtla.2022.101362.

Aavash Budhathoki

Tungsten (W) exhibits exceptional properties, including high density, melting point, thermal conductivity, stiffness, and strength, making it an ideal candidate for extreme environments. However, its limited ductility at room temperature poses significant challenges for fabricating complex geometries using conventional manufacturing methods. Additive manufacturing (AM) provides a pathway for producing intricate tungsten components, but the high energy densities required for melting result in slow processing rates. To address this limitation, a hybrid manufacturing approach combining AM with hot isostatic pressing (HIP) is explored. HIP enables the full densification of tungsten by applying high pressure and temperature over controlled timescales, effectively eliminating porosity and enhancing mechanical properties. This study integrates finite element modeling (FEM) to optimize the HIP process, focusing on plasticity, creep, and powder densification models essential for accurately simulating material deformation and shrinkage. The models incorporate temperature- and pressure-dependent behavior to predict densification kinetics and final component geometry. By leveraging simulation-driven design, the study aims to establish an efficient methodology for producing fully dense tungsten components with optimized geometries, overcoming the inherent limitations of AM and traditional powder metallurgy techniques.

Computational Materials Science

Image by Itamblyn CC BY-SA 3.0

Sean Lam

Developing Quantum Machine Learning Techniques to Enhance Microscopy Data Analysis

Sean Lam ^1,2, Roberto dos Reis ^3,4,5

¹ Department of Physics, Colorado College, Colorado Springs, CO
² Department of Chemistry & Biochemistry, Colorado College, Colorado Springs, CO
³Department of Materials Science and Engineering, Northwestern University, Evanston, IL
⁴ The NUANCE Center, Northwestern University, Evanston, IL
⁵ International Institute of Nanotechnology, Northwestern University, Evanston, IL

Advancements in quantum computing present new opportunities for enhancing microscopy data analysis and simulation. This project explores the integration of quantum algorithms to improve the efficiency and precision of electron microscopy data processing. Our workflow consists of data preparation, quantum state encoding through Pauli feature maps, application of parameterized quantum circuits, and measurement-based decoding to retrieve and visualize microscopy features. We utilize abTEM package for simulating electron microscopy data, generating controlled datasets for algorithm benchmarking. Future work will focus on refining quantum filtering techniques, training on larger datasets, and advancing feature detection capabilities. This research contributes to the growing intersection of quantum computing and computational microscopy, with the potential to reduce computational overhead for complex data analysis while increasing feature detection sensitivity in materials science applications.