Massachusetts Institute Of Technology

NSF Awards · FY 2025 · 2025-12

Growth is a quintessential feature of all living systems; understanding the mechanics of growth is crucial in a wide range of ecological, industrial, and medical settings. However, while there is an increasing appreciation for the significant role of mechanics in defining the growth and form of biological materials, the field has yet to provide a basic understanding of key mechano-morphogenesis processes and their sensitivity to various environmental factors, such as geometrical constraints and nutrient availability. To address this question, this collaborative project takes advantage of a highly tunable biological system that is capable of macroscale growth - bacterial biofilms. Confocal imaging and analysis of the growth process looks to enable detailed observation of various growth phenomena at both single-cell and continuum levels and can measure the influence of environmental factors. The parallel theoretical effort will stem from the derivation of theoretical models that integrate only the essential ingredients by which the biological system evolves seeking to provide an open-ended strategy to explore and expose rules and unexpected phenomena in morphogenesis. The research is likely to have direct implications for our understanding of the development and resilience of bacterial biofilms. The overarching goal of this collaborative research is two-fold: 1) to deepen the understanding of the development of biofilms in constrained environments; 2) to leverage the growth of highly tunable biofilm systems as a generic scheme for biological growth. The approach focuses on the development of theoretical models that are complex enough to contain the essential coupled mechanisms involved in growth and morphogenesis but are simple enough to explain the basic phenomena that may emerge and can serve as tools to expose additional unexpected phenomena. The first two objectives of this work study the separate roles of nutrient transport and mechanical stress using specially designed experimental setups that isolate the specific phenomena of interest in the embedded biofilm system. The third objective further iterates between the theory and the experiments by looking to capture the coupling between the different mechanisms and to explore ranges of response that are beyond reach of the experimental system. The developed models and the conclusions obtained from observations of bacterial biofilms system confined in hydrogels look to be applied to other cellular collectives or biological entities growing under mechanical constraints. These insights could be applied to several biomedical applications and potentially open new directions for studies on embedded biofilms, such as their antibiotic resistance. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-11

PROJECT SUMMARY/ABSTRACT The pharmaceutical and agrochemical industries rely heavily on the stereoselective preparation of chiral molecules because opposite enantiomers often demonstrate disparate activities. Though numerous approaches exist to obtain enantiomerically enriched material, these methods typically require the synthesis of specialized starting materials/reagents. The catalytic deracemization of racemic material is an attractive alternative, as racemic synthesis is often straightforward and enantioinduction can be achieved by a catalytic rather than stoichiometric process. In recent years, photocatalysis has emerged as a strategy for driving systems out of equilibrium and accomplishing deracemization reactions. However, despite the prevalence of the α-carbonyl stereocenter in peptidomimetics, natural products, and pharmaceuticals, the catalytic deracemization of this scaffold has not been achieved in a general sense; this may be in part due to the challenges associated with generating a reactive intermediate at the acidic α-position under mild conditions, coupled with the lack of nucleophilic HAA reagents. This proposal seeks to unlock a new reactivity platform within the field of stereochemical editing and utilize it to enable deracemization of α-carbonyl stereocenters. Leveraging the Wendlandt group’s expertise in asymmetric catalysis, we will investigate stereoselective hydrogen atom abstraction (HAA) using chiral amine-borane catalysts and develop a deracemization protocol for the conversion of racemic starting materials into high-value enantioenriched products. Prior literature has established proof of concept for kinetic resolution of α-carbonyl stereocenters with these reagents, albeit with modest selectivity outcomes; however, their extension to deracemization has not yet been investigated or accomplished. We seek to leverage these chiral amine-boryl radicals and photoredox catalysis for the development of a novel method for the deracemization of α-carbonyl stereocenters, with a particular focus on biologically active motifs such as β2- and γ2- amino acids. Ultimately, we seek to establish amine-boranes as a modern reagent class for selective HAA of acidic C–H bonds and gain a deeper understanding of the factors affecting enantioselectivity of HAA with amine-boryl radical species. We will accomplish these goals through synthesis of novel catalysts and evaluation of their reactivity via spectroscopic studies. Research will be carried out at MIT, an institution that excels in synthetic chemistry. I will receive training in asymmetric catalysis from Prof. Alison Wendlandt, a trailblazer in the stereochemical editing field, training in photophysics from Prof. Schlau-Cohen, an expert in ultrafast spectroscopy, as well as training in organoboron chemistry through informal interactions with Prof. Robert Gilliard and his research group at MIT. The proposed research is highly complementary to my doctoral training in the data science-driven study of pharmaceutically relevant reactions, enabling me to expand my skillset to include asymmetric catalysis, photophysics, and main group catalysis.

NSF Awards · FY 2025 · 2025-11

Today's frontier Heliophysics research topics require fusing data from multiple distinct sources in order to advance community knowledge. However, creating analysis-ready data from multiple data sources remains burdensome. A more uniform way to discover measurements ripe for synthesis, as well as a consistent and efficient method to subsequently obtain the data in a common analysis environment for the study in question, is urgently needed to produce future discoveries while following interoperable, reusable principles and open science practices. This project provides concrete, usable, and valuable progress by linking the NSF-sponsored Madrigal Database, responsible for storing and providing access to a vast variety of ground-based geoscience data, with two emerging tools for standardizing data access, organization, and metadata: (1) the Heliophysics Application Programming Interface (HAPI), a time series download service and a tool bundle; and (2) the Space Physics Archive Search and Extract (SPASE) data model. This project will integrate the NSF-funded Madrigal, the database for upper atmosphere science, with the Heliophysics Application Program Interface (HAPI) and Space Physics Archive Search and Extract (SPASE) model. The project will increase the Findable, Accessible, Interoperable, and Reusable (FAIR) of the data and enable scientific discovery across a wide range of heliophysics with five major activities: 1) development of a Madrigal-HAPI server to serve Madrigal data via HAPI infrastructure, 2) development of a metadata converter between HAPI and SPASE, 3) development of a Madrigal-specific plugin to populate SPASE metadata, 4) research community engagement and training, and 5) quantitative assessment of the efficacy of newly developed tools. This award by the Office of Advanced Cyberinfrastructure in the Directorate for Computer and Information Science and Engineering is jointly supported by the Directorate for Geosciences. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

This project aims to serve the national interest by expanding access to high-quality, adaptable STEM learning materials through the development and large-scale deployment of Open Educational Resources (OER) and integrated social annotation technologies. This Level 2 Institutional and Community Transformation project addresses the ongoing challenge of limited access to affordable and pedagogically effective instructional materials, particularly in large educational systems. At the core of the project is the enhancement and deployment of Nota Bene (NB), a web-based annotation platform that enables students and instructors to engage in contextual conversations directly alongside the textbook content - the digital version of marking up physical textbooks. By analyzing these in-line comments, instructors can identify confusing, outdated, or incomplete material, and students can contribute insights or suggest improvements that reflect their learning needs. This project will integrate Nota Bene into the LibreTexts platform, one of the most widely used OER repositories, with the aim of transforming passive reading into an interactive and collaborative STEM learning experience. The project's goals are to establish social annotation as a core practice in STEM education, to create a scalable infrastructure for collaborative OER development, and to generate actionable knowledge on how learner-instructor interactions through annotation drive content improvement and learning outcomes. Specifically, the project will (1) redesign and expand the NB platform to enhance usability for different types of STEM courses and institutions nationwide, (2) develop and implement workflows for peer review and open pedagogy that embed continuous feedback loops into OER lifecycle management, and (3) build and operationalize CalOPEN, a statewide hub supporting OER creation, adoption, and innovation within California's higher education systems, serving as a replicable national model. The project employs design-based research to iteratively improve the Nota Bene platform and its pedagogical integration, coupled with large-scale learning analytics to examine patterns in annotation use, feedback quality, and their correlation with student learning gains. Assessment and evaluation will be multi-faceted, involving external evaluators who will conduct surveys, interviews, system usage analysis, and performance assessments to measure impact on student engagement, instructor practices, and content quality. The NSF IUSE: EDU Program supports research and development projects to improve the effectiveness of STEM education for all students. Through the Institutional and Community Transformation track, the program supports efforts to transform and improve STEM education across institutions of higher education and disciplinary communities. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Social scientists and decision makers have been interested in whether and how well they can generalize the results of randomized controlled trials (RCTs) to new populations, places, and contexts. This question of external validity is a foundation of evidence-based decision making because study populations and sites are often different from the real-world populations and sites to which an intervention might be scaled up. This research develops statistical approaches to improve the external validity of RCTs in the social sciences. The research makes two methodological contributions. The first is a new framework to design RCTs for external validity. Specifically, this project develops an algorithm to select experimental sites such that researchers can credibly estimate generalizable causal effects. Site selection is essential because experimental results in the social sciences are often heterogeneous across places, and biased selection of experimental sites can lead to low generalizability and replicability. The second contribution is a statistical tool to quantify the robustness to external validity bias. This new measure of external robustness allows researchers to evaluate external validity even when RCTs were conducted without external validity considerations. These two methods complement each other and together provide a unified pipeline to improve the external validity of RCTs in the social sciences. The project is co-funded by the Science of Science: Discovery, Communications, and Impact program; the Accountability, Institutions and Behavior program; and the Office of Integrative Activities. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Many distributed systems involve interactions among computers controlled by different parties. These systems must work correctly even when some of the participants are malicious and try to interfere with the system. In computer science , these kinds of malicious participants are called Byzantine participants. Various protocols and algorithms have been developed to ensure that the systems operate correctly so long as only a small fraction of the participants behave maliciously. However, implementations of these kinds of systems often have bugs. One class of bugs that is particularly challenging to address is liveness bugs, in which the system stops making progress and fails to complete operations. To reduce the incidence of bugs, researchers have developed an approach called formal verification, in which a mathematical proof is constructed that shows a software system is free from a certain class of bugs. However, existing methods for verifying the absence of liveness bugs have limitations that make them inapplicable to many important systems. This project develops new techniques for verifying the absence of liveness bugs in systems with malicious participants, expanding the kinds of systems that can be verified. In addition, the research team develops new tutorials, labs, and lectures on verification of distributed systems, and organizes the annual New England Systems Verification Day, which brings together verification researchers and industry practitioners. This project develops a new automation-friendly approach for modularly verifying the liveness of distributed systems with Byzantine participants. Especially, the project team develops novel liveness-preserving composition operators, which allow decomposing complex system proofs into smaller independent subprotocols. To reason about the use of cryptographic signatures in the distributed systems that tolerate Byzantine participants, the project uses a new model called the stapling model, which enables to analyze the signed messages in an automatable way. To prove the soundness of the stapling model and other components of the approach, this project develops an innovative program logic for proving liveness based on separation logic. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

This award supports the 19th Graduate Climate Conference (GCC), planned for 7-9 November 2025 at the Marine Biological Laboratory at Woods Hole, Massachusetts. The GCC is an interdisciplinary climate conference organized entirely by and for graduate students. It provides a discussion forum for students conducting research on climate in a variety of disciplines including atmospheric and oceanic dynamics, biogeochemical cycles, clouds and aerosols, and the cryosphere. The meeting is intended to help students familiarize themselves with the breadth of climate science, gain exposure to the enormous range of tools available to address climate-related questions, and understand how their research fits into the broader landscape of current climate science. The conference seeks to better prepare graduate students for scientific inquiry in a field that increasingly demands interdisciplinary approaches. Funds provided here support travel and subsistence for 120 students chosen through a competitive process. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

This award supports one year of continued operations of the Event Horizon Telescope (EHT) as a facility for frontier science, leveraging extensive NSF and international investment. Initially conceived as an instrument to produce images of supermassive black holes (SMBH), recent upgrades allow the EHT to explore the dynamics of accretion and outflow around black holes using time-lapse monitoring. New operating modes that will increase open access to the EHT will be tested. EHT has seen significant increases in capability by deploying new telescopes and receivers, developing advanced analysis and imaging algorithms, and enhancing observing agility. The EHT Collaboration (EHTC) also provides analysis and imaging tools and calibration information to help both key black-hole dynamics studies and PI-led projects. Training of junior scientists is foundational to the EHTC, as they have essential roles in telescope operations, data calibration, and knowledge generation. Undergraduates work on related projects. The public response to the first shadow images shows their enduring fascination with black holes and the extraordinary outreach potency of the EHT. Public dissemination and outreach will continue. The EHT is a network of (sub)millimeter telescopes that uses the technique of very long baseline interferometry (VLBI) to achieve the sharpest angular resolution of any current astronomical instrument. Primary science drivers include determining the origin of relativistic jets, testing the spin extraction hypothesis, understanding flares, and characterizing the population of SMBH in galactic nuclei. Dynamical multi-wavelength observations of the black hole in our Galactic center (Sgr A*) will probe flares and address why Sgr A* exhibits weaker variability than predicted by relativistic modeling. New spectral-line VLBI capabilities will enable studies of submillimeter masers and molecular absorption lines at higher resolution than presently possible. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Modern society depends on analyzing massive amounts of personal information to improve healthcare, enhance national security, and drive economic growth. However, current practices for storing and sharing sensitive data have led to major data breaches exposing millions of people's personal information, including medical records, financial details, and government secrets, undermining public trust and threatening national security. This project addresses this critical challenge by developing new computer systems that allow organizations to gain insights from large datasets while keeping individual information completely private. This project will bring privacy protections to each of the three steps of the data-management lifecycle: data collection, data processing, and data retrieval. By protecting privacy during data collection, analysis, and retrieval, this research serves the national interest by enabling continued technological advancement while safeguarding citizen privacy, supporting economic competitiveness in data-driven industries, and strengthening cybersecurity infrastructure. This project investigates three fundamental research areas to advance privacy-preserving data systems. First, the research team will develop new protocols for privacy-preserving data collection that enable servers to compute aggregate statistics over client data without accessing individual records, with emphasis on reducing computational costs and expanding the class of computable functions compared to existing systems. Second, the project will design privacy-preserving machine learning algorithms for training recommender systems, clustering algorithms, and decision trees that operate on encrypted data while maintaining model accuracy. The third and last component of the project will be to develop new techniques that let clients privately query server-side datasets. In this thrust, the project will develop a relational database that supports private queries. A key design component will be new data structures, optimized to work with cryptographic privacy-protecting protocols. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

This project will advance our capabilities to understand and predict two critical aspects of the Earth system by pioneering a new approach driven by artificial intelligence (AI). For decades, the complexity of the underlying physics has limited progress in quantifying the cooling effect due to atmospheric aerosols and their effect on clouds and the rate of Arctic sea ice melt. This project tackles these challenges by using advanced AI to learn directly from satellite observations and laboratory data, developing more accurate and reliable computer simulations. These will in turn result in improved predictive capabilities that are vital for U.S. strategic interests, for example, as a warming Arctic opens new maritime shipping routes essential for commerce and security. More reliable environmental intelligence will support better-informed decisions for infrastructure planning and risk assessment. The project will also make all its AI tools and software openly available and will train a new generation of researchers in these cutting-edge methods. To address current limitations in Earth System Models (ESMs), this project will develop and implement novel parameterizations for aerosol-cloud interactions (ACI) and Arctic sea ice thermodynamics. The research will leverage AI, specifically a novel framework called Ensemble Kalman Diffusion Guidance (EnKG), to learn from a diverse range of observational and laboratory data. For ACI, the project will develop new data-driven models for how aerosols form cloud droplets and ice crystals, using high-fidelity simulations for pre-training before online fine-tuning within the ESM developed by CliMA, the Climate Modeling Alliance. For sea ice, the research will build an improved thermodynamics model, incorporating machine learning components to better represent processes such as melt ponds and albedo feedback. The EnKG framework will be developed to efficiently train these embedded ML parameterizations using large-scale satellite observations without requiring model derivatives. Finally, the project will conduct ESM simulations using the new parameterizations to provide improved, uncertainty-quantified estimates of aerosol radiative forcing and more robust projections of future Arctic sea ice decline. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Non-technical description Next-generation low-power electronics, wireless technologies, and the Internet of Things (IoT) require diode devices that can operate efficiently at high frequencies, with low power input and minimal energy loss—capabilities that remain challenging for conventional diode systems. This project focuses on a new class of materials called quantum diodic magnets (QuDiM), which host “quantum dipoles”—dipolar distributions of quantum wavefunction properties that enable current rectification, where electric current flows more easily in one direction than the other, a defining characteristic of diode function. Unlike traditional diode materials, QuDiM systems can operate at very high frequencies and low power with minimal energy loss. Importantly, their performance is resilient to impurities and thermal fluctuations. This intrinsic robustness reduces the need for ultra-clean materials, simplifies device design, and makes these systems potentially suitable for diverse environments. To accelerate progress in this emerging field, the research will build an integrated discovery pipeline linking theory, computation, synthesis, and experimental characterization. In parallel, this project will promote interdisciplinary education and open science by developing teaching modules that introduce students to computational, data-driven, and AI-based approaches in quantum materials research. Technical description This project aims to design, synthesize, characterize, and benchmark quantum diodic magnets (QuDiM)—a class of quantum magnetic materials that exhibit intrinsic nonreciprocal transport, enabling direction-dependent conductivity for electrical and microwave rectification. This emerging diode technology is rooted in geometric quantum properties, such as Berry curvature and quantum metric dipoles. Unlike conventional diodes, QuDiM materials can achieve efficient high-frequency rectification in the ultralow-power regime—a performance space previously inaccessible with traditional mechanisms. Crucially, their nonreciprocal response is dissipationless, remaining robust against impurity scattering and electron-phonon interactions. This intrinsic resilience enables functionality at elevated temperatures, including above room temperature, while reducing the need for high-purity materials and simplifying device and circuit design. The development of such quantum dipole-enabled materials remains in its early stages. To accelerate progress, the project brings together an interdisciplinary and international team with expertise in quantum theory, high-throughput computation, machine learning, multi-route chemical synthesis, and advanced experimental characterization. A co-design framework integrates theoretical modeling, computational screening, and experimental realization through iterative feedback between design and measurement. Automation and AI-driven strategies will further accelerate the discovery and optimization of QuDiM systems, with the overarching goal of establishing a robust materials platform for next-generation low-power electronic and wireless technologies. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

In this ASCENT project, the team aims to develop a set of semiconductor technologies, including new device fabrication, large-scale heterogeneous integration, and robust beam-alignment system architectures, to achieve electrically controlled collimation of electromagnetic waves in sub-terahertz (sub-THz) frequency bands at low cost. Compared to the existing 5G wireless bands below 100 GHz, the sub-THz bands offer unprecedented wide bandwidth that can enable ultra-fast data rate for data center networking and wireless infrastructure, as well as high precision for radar and imaging. The electronic hardware developed in this project provides a highly desirable function for most sub-THz systems -- focusing the beam power within one degree in space (hence the term "needle beam" in the project title) with high-precision electronic control of the beam direction. The new hardware architecture enables wireless communication systems to achieve a high data rate up to 120 Gbps over a distance greater than one kilometer. It also enables radar imaging systems to achieve high-resolution sensing of the ambient environment, which is critical for all-weather safe operation of autonomous vehicles. In addition, this project not only provides extensive graduate researcher training in high-frequency circuit designs and advanced semiconductor manufacturing but also promotes STEM education through various programs. Needle beam forming at 140 GHz requires large (> 70x70 millimeter square) electronic phase-controlling surfaces with low signal loss at high frequencies. Therefore, transistors fabricated using advanced lithography are needed. Recent demonstrations of sub-THz reflectarray using either tiled complementary metal-oxide-semiconductor (CMOS) FinFET chips or chip package modules have excessively high fabrication and assembly costs. In this project, the team explores a new direct wafer-scale integration approach through low-temperature fabrications of custom metal-insulator-semiconductor-insulator-metal (MISIM) variable capacitance devices and high-efficiency sub-THz antennas on top of a foundry-processed integrated-circuit glass substrate. Without using any expensive advanced lithography, the devices can still achieve low-loss phase shifting of sub-THz signals, hence significantly reducing the cost of the needle-beam-forming system. The project also investigates new reflectarray circuits that perform self-correction of device defects and process variations, as well as new reflectarray transceiver architectures that enable compact overall system form factor and high-precision beam alignment. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

This project will build the first multiscale AI ocean emulator, spanning from submesoscales to large-scale ocean flows, for long-term ocean variability. The emulator will be used to investigate the complex nonlinear dynamics of the ocean circulation and to bridge the gap between understanding and simulating ocean variability across many time and space scales. This project will also advance the development of physics-informed, autoregressive neural networks capable of learning from heterogeneous, multi-resolution datasets. It addresses core challenges in scientific machine learning, including stability and interpretability in complex systems, and introduces methodologies that can extend to other multiscale problems. Understanding the interactions between oceanic processes across scales is essential for advancing our knowledge of ocean circulation, heat and momentum transport, and their role in shaping long-term variability. The primary objective is to investigate the spatio-temporal interactions between submesoscales, mesoscales, and large-scale flows on regional and global scales. This will be achieved by building a three-dimensional multiscale AI emulator, using deep neural networks, trained on a suite of numerical and observational datasets at different spatio-temporal resolutions. The questions to be addressed include: How to construct a physically-based 3D AI ocean emulator from heterogeneous datasets? How to evaluate a multiscale emulator from sparse and imperfect datasets? What fraction of ocean submesoscale and mesoscale physics drives momentum, energy, and heat transport at large scales? What is the role of multiscale processes on local ocean variability, such as marine heat waves and sea level extremes? This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Breast cancer is the most commonly diagnosed cancer among women in the United States, yet many individuals face barriers to routine screening due to limited access, high costs, and discomfort. While mammography is the clinical standard, it is radiative and less effective for younger women and those with dense breast tissue. In response, the U.S. Food and Drug Administration now requires that patients be notified about their breast density and advised that additional imaging methods may improve cancer detection. Some aggressive tumors—known as interval cancers—can develop between mammograms and progress rapidly, making early detection especially important. Ultrasound is a safe, noninvasive, and accessible tool for breast cancer screening. However, current handheld ultrasound (HHUS) imaging is highly operator dependent, requires significant training, and typically offers only two-dimensional (2D) imaging, making it easy to miss anomalies unless the imaging plane intersects them precisely. Automated Breast Ultrasound (ABUS) systems provide standardized three-dimensional (3D) imaging with reduced operator variability but are expensive, clinic-bound, and reliant on uncomfortable compression and mechanical scanning. To overcome these limitations, this project introduces a wearable, real-time 3D ultrasound system that is low-power, and optimized for wide-angle, high-resolution volumetric imaging. The system is intended as an adjunct to routine mammography, designed to improve access to early detection, making breast imaging more accessible and frequent outside of clinical settings—particularly for individuals with dense breast tissue or limited imaging access. The system will be evaluated in clinical studies to assess diagnostic performance and generate datasets for developing AI tools to assist in early anomaly detection, which aligns with NSF’s mission to improve national health outcomes. The project includes educational and outreach efforts to spark interest in STEM at the K–12, undergraduate, and graduate levels. Novel fabrication techniques and wearable imaging concepts will be shared through new course modules and public demonstrations, aiming to engage students and educators beyond traditional physics and materials science disciplines. This project proposes a fully integrated, wearable ultrasound system to enable autonomous, longitudinal breast health monitoring. It combines innovations in patch design, transducer fabrication, miniaturized electronics, efficient three-dimensional (3D) imaging on curved anatomy, and AI-driven analysis to deliver safe, accessible, and gel-free imaging in a portable format. We introduce new laser-based micromachining strategies for piezoceramic dicing and electrode patterning, which will overcome limitations of conventional planar fabrication techniques such as dice-and-fill. These methods reduce mechanical stress, improve yield, and streamline the manufacturing process for 2D transducer arrays. Newly engineered matching and backing layers will improve bandwidth, enhance acoustic coupling, and suppress transducer ringing. Combined with biocompatible dry coupling materials, this approach eliminates the need for ultrasound gel and enables conformal, reusable skin contact. A soft substrate designed to have minimal-stress on the component will facilitate conformable and repeatable probe positioning, while a ring-array architecture will enable acoustic triangulation, allowing accurate 3D image reconstruction over complex curvilinear targets such as the breast. A novel signal acquisition architecture, “Chirped Data Acquisition System (cDAQ),” will be implemented to achieve high signal-to-noise ratio at sub-Nyquist sampling rates. In parallel, a “Convolutional Optimally Degenerate Array (CODA)” will be developed to replicate full-matrix resolution with significantly fewer transducer elements. In vivo trials with Massachusetts General Hospital will assess diagnostic performance, and the data will also be used to train AI algorithms for real-time detection and classification of breast anomalies. This work advances ultrasound transducer manufacturing, portable low-power imaging systems, and scalable clinical validation, providing a foundation for broader deployment of breast health screening technologies. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

From online advertising to content delivery networks and recommendation systems, many modern technologies rely on algorithms that must operate in real time without knowing what will happen next. Researchers in computer science, operations research, engineering, and other fields have developed powerful online optimization and learning tools for making effective decisions in the face of uncertainty, by helping systems learn from past outcomes to improve performance over time. However, these methods are vulnerable to attackers aiming to disrupt the system: fake reviews, ad fraud, denial-of-service attacks, and other attacks can corrupt the algorithms' learning process. Researchers have begun developing "corruption-robust" learning algorithms that are more resilient to attacks; however, significant barriers remain in translating these theoretical advances into real-world systems. This project aims to reduce those barriers by designing learning algorithms that are both theoretically sound and practical to implement, enabling more robust decision-making in real-world applications. This work directly supports the national interest by strengthening the resilience of critical cyberinfrastructure and advancing the scientific foundations of trustworthy AI. This project advances the theoretical foundations of online optimization and learning by explicitly incorporating robustness to security threats and adversarial corruptions. While online learning has been extensively studied for decades across computer science, control theory, economics, and operations research, a rigorous understanding of its vulnerabilities and resilience under malicious attacks remains underdeveloped. This project fills this gap by developing new corruption-robust algorithms for sequential decision-making problems, particularly in settings involving combinatorial action spaces and resource constraints in both single-agent and multi-agent learning environments, addressing various classes of adversarial threats. The research involves designing algorithms with provable performance guarantees in the presence of corruption, analyzing their theoretical properties, and developing efficient implementations. To validate its impact, the project evaluates the proposed methods in two real-world applications: probabilistic maximum coverage for content delivery networks and online learning to rank in social and e-commerce platforms. The project's outcomes will contribute broadly to secure and trustworthy decision-making systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Surface waves and their resulting turbulent processes play a central role in regulating the exchanges of mass, momentum, and energy between the atmosphere and ocean, directly influencing sea states, weather patterns, and climate systems. Although integrating the dynamics of surface waves into the coupled atmospheric-oceanic models is crucial for accurate weather and climate forecasts, the current understanding of fundamental processes that link the turbulent flow structures above and below the surface within the coupled air-sea boundary layers is limited. This is due to challenges in resolving the dynamics of small-scale turbulence in the vicinity of the air-sea interface. This project will combine high-resolution laboratory experiments, high-fidelity numerical simulations, and foundational AI techniques to examine the multiscale turbulence above and below ocean surface waves and quantify the two-way coupled air-sea momentum and energy fluxes at the air-sea interface. This project will examine the coupling of wave-induced flow structures above and below the air-sea interface. A synergistic experimental and AI-driven approach will be used to develop a comprehensive parameterization of wave and turbulent stresses at the air-sea interface. The aim is to address the turbulent closure problem in the governing equations and improve predictive models of air-sea fluxes. A combination of novel experiments that concurrently measure air- and water-side flow velocities and an advanced multiscale AI-driven framework that completes the partially measured statistical signatures of the flow will be employed. The AI model will integrate attention mechanisms with physics-informed neural networks (PINNs) to enhance the two-dimensional planar velocity measurements by reconstructing the third velocity component and the pressure field. The resulting dataset will support the development of data-driven surrogate models for air-sea fluxes with enhanced physical consistency and superior generalizability. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Non-technical Description: The development of transformative, next-generation quantum devices will require breakthrough innovation in molecular materials. Nature has evolved the sophisticated ability to organize chromophores and pigments to harness quantum mechanical properties for powerful capabilities, including solar energy harvesting, nanoscale energy transport, and energy conversion. Replicating these quantum capabilities of nature using synthetic materials remains a significant goal of rational materials design and fabrication. To overcome the bottlenecks that are intrinsic to modern materials design efforts, including costly and time-consuming synthesis and experimental characterization, the project will develop a machine learning-guided, high-throughput in silico screening platform to accelerate the discovery of functional chromophore- and qubit-DNA assemblies. The research team seeks to achieve predictive control over electronic state evolution in molecular systems organized using designer DNA assemblies. Realizing this control will enable the design of revolutionary materials and devices for various applications in quantum science and technology, including next-generation photovoltaic coatings, quantum sensors, quantum simulators, and biological imaging agents. Technical Description: DNA origami has the ability to precisely position small molecule geometries and controllably modify their local molecular environments. This ability can be leveraged to endow materials with tailored electronic properties. The project aims at designing nanomaterials that (1) extend the lifetime of optical excitations for application to solar energy conversion and storage and (2) protect and maintain coherence in molecular spin-center arrays for application to quantum sensing and computing. The originality in the research approach lies in the fact that this leverage will be used to explicitly and independently target and manipulate the properties of the molecular system’s environment (i.e., bath and system-bath interactions) to protect and enhance quantum properties. The project will develop an open-source platform for computational high-throughput screening of DNA-small molecular hybrid nanostructures that can be used to accelerate the development of materials with novel and controllable electronic properties. The project will provide cross-disciplinary training for graduate and undergraduate students to learn innovative methods that span theory, computation, DNA synthesis, and spectroscopic characterization. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Building on fundamental innovations in the study of the intersection of chemistry and data science, the Community Resource for Innovation in Polymer Technology (CRIPT) project built a stable and scalable data ecosystem that makes polymer data widely available to and usable by a broad community. The current project, the Community Resource for Innovation in Materials and Chemistry (CRIMaC), substantially enhances CRIPT to grow beyond the community of 160,000 polymer researchers into a tool that potentially enables millions of innovators in chemistry and materials research to find data faster, collaborate more easily, and to use trustworthy data to drive innovations in artificial intelligence, machine learning, and automation. CRIMaC includes innovations to address user needs for visualization and data reliability, and its functional scope is broadened through extensions of innovations in data science originally developed for polymers to be relevant to chemical and materials science data more broadly. The technical innovations and cyberinfrastructure operations are coupled with educational and outreach efforts to grow the CRIMaC user community, including curricula material for undergraduates, online tutorials for self-learning, and train-the-trainer opportunities for advanced users and instructors. CRIPT is a stable, scalable, and robust platform for polymer data based upon key innovations in data structures, molecular representations, search technologies, and similarity rankings that enable cheminformatics beyond small molecules. This project transforms CRIPT into a more general cyberinfrastructure called CRIMaC through three key pillars of cyberinfrastructure development that substantially enhance functionality and broaden CRIMaC’s reach beyond polymer science. With improved interactive data visualization, CRIMaC will transform CRIPT data structure to develop a novel concept of “graph as comment” using the Vega-Lite notation, enabling domain experts to drive conversations about data based on graphical objects. To ensure data trustworthiness (a major challenge for data systems based on user contributions), CRIMaC designs a 4-pronged validation scheme using rules-based validation, transparent data provenance, AI-driven outlier detection, and crowdsourced validation. Moving beyond polymer science requires the innovation of new cheminformatics tools that extend CRIMaC’s foundational BigSMILES technologies into biochemistry, theory/simulation, and organic/inorganic hybrid materials, capturing the full range of chemistry and soft materials. To support and expand adoption of CRIMaC, the project also teaches the next generation of knowledge workers to use these types of data systems using an approach that a) develops educational materials to teach informatics and data systems directly in undergraduate curricula, b) provides tutorials to allow users to self-learn the CRIMaC system, and c) implements a train-the-trainers course that enables individuals to gain a marketable credential in the use of CRIMaC data system as a tool. This approach simultaneously promotes the uptake of CRIMaC, encourages super-users of the open-source ecosystems, and provides professional development to the community of practitioners. This award by the Office of Advanced Cyberinfrastructure is jointly supported by the Division of Materials Research and the Office of Strategic Initiatives in the Directorate for Mathematical and Physical Sciences. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY/ABSTRACT Maintaining the equilibrium of the various ribonucleoside and deoxyribonucleoside triphosphates (NTPs and dNTPs) within cells is critical for homeostasis. As the only de novo mechanism for dNTP synthesis, ribonucleotide reductases (RNRs) are essential enzymes in all forms of life. These metalloenzymes catalyze the reduction of ribonucleotides into deoxyribonucleotides, which is essential for maintaining the dNTP pools for DNA synthesis and repair. Dysregulation of RNR activity is associated with tumorigenesis, and cancer cells are overly reliant on the enzyme as a result of their rampant proliferation. Therefore, inhibitors of RNR proteins are effective anti-cancer therapeutics and have been proposed as potential antibiotics. However, to date, there is no structure of human RNR in the active state, and many aspects of the protein’s regulation are still unknown. Recent structural work on RNR proteins has resulted in high-resolution structures of the active complex of the E. coli enzyme, which is a class Ia RNR like the human enzyme. These enzymes consist of two dimeric subunits referred to as α2 and β2. Although these enzymes are similar, there are also significant differences between the species. For example, while both proteins form rings when inactivated, the content of these rings are different, and preliminary data indicates that the inactivation mechanism is also distinct. The goal of this project is to use cryogenic electron microscopy (cryo-EM) to determine structures of the human RNR protein in both active and inactive states. My first aim is to determine cryo-EM structures of the α2 subunit bound to both ATP and dATP in the allosteric activity effector pocket within the regulatory domain. These data will reveal any conformational changes of the regulatory domain due to the differential binding and ATP and dATP that could not be observed in the crystal structures due to lattice contacts. I will then determine the structure of the human enzyme in the active α2β2 conformation to identify key features and interactions within the human enzyme. Together, these structures will uncover details about this integral protein and how it functions, which I will interrogate using site-directed mutagenesis and biochemical assays. Our findings will aid in the development of novel RNR inhibitors for the treatment of cancer or bacterial infections. As one of the world’s most well-renowned research institutions, MIT is an ideal environment to complete these aims. My mentor, Dr. Catherine Drennan, is an international leader in the field of structural biology with a focus on the study of metalloenzymes. Her mentorship will enable me to excel in my postdoctoral research and allow me to reach my best as a scientist moving forward.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY/ABSTRACT Acute respiratory distress syndrome (ARDS) is a severe form of lung injury with significant public health implications due to severe morbidity and mortality. The need to utilize existing data to inform prospective research and clinical decision making was emphasized during the COVID pandemic, when ARDS became a leading cause of death, and clinicians were forced to operate outside of existing evidence. Many commonly applied interventions including use of neuromuscular blockade (NMB), steroids, driving pressure and mechanical power are used despite negative data, without high quality prospective studies or with equivocal evidence and with the potential for both benefit or harm. Additionally, phenotypes of ARDS may have different prognosis and response to treatment, but thus far have not been well differentiated using routinely available dynamic clinical data, nor have they been incorporated into prospective trials. ‘Dynamic treatment regimes’ (DTRs) are rules for making treatment decisions sequentially at multiple time-points based on a patient’s evolving history. Most relevant treatment strategies for ARDS are DTRs. DTRs may be evaluated in randomized trials, however it is infeasible to conduct randomized trials testing all DTRs of interest. This grant proposes ‘target trial emulation’ from observational data using ‘g-methods’ for confounding adjustment to address multiple gaps in our knowledge about ARDS care. We will address important methodological gaps in ARDS phenotyping and develop advanced machine learning (ML) methods for dynamic phenotyping for prognostication and personalized DTRs to determine for whom and when specific ARDS treatments are beneficial. The investigation will utilize three large datasets—including the Medical Information Mart for Intensive Care (MIMIC) IV database, the eICU collaborative research database, and the Dutch AmsterdamUMCdb database—representing a wide geographic and demographic spectrum, and the ability to assess stability of findings across geography and centers. To address these knowledge gaps regarding use of NMB, steroids, driving pressure and mechanical power, as well as identify phenotypes of patients most responsive to treatment, we propose two overarching aims for this grant. Specific Aim 1) Using target trial emulations and g-methods, we will estimate clinical outcomes that would result under a range of treatment strategies for NMB, steroids as well as driving pressure and mechanical power thresholds. Specific Aim 2) We will develop machine learning methods to derive dynamic markers for ARDS phenotyping and formulate personalized DTRs for ARDS treatment. The project represents a collaborative effort between experts in critical care medicine (with a specialty in mechanical ventilation), machine learning, and causal inference. Our results will address important gaps in clinical knowledge about treatment of ARDS and inform the design of future randomized trials. Our study designs, code, and constructed cohorts will also provide valuable templates for other researchers to use in future observational studies, which we foresee will broadly improve the quality of evidence from observational data in critical care.

NSF Awards · FY 2025 · 2025-09

This I-Corps project focuses on the development of shelf-stable, ready-to-use 3D bioprinted tissue models for preclinical drug development. Current in vitro models, such as 2D cultures, spheroids, and organoids, are useful in certain contexts but the structural and cellular complexity need to more accurately replicate human physiology. As a result, the printed tissue models often fail to predict clinical outcomes, contributing to high drug attrition rates. Regulatory trends increasingly favor alternatives to animal models, highlighting the urgent need for advanced, human-relevant in vitro systems. While 3D bioprinting offers a path forward, its impact has been limited by the inability to store, manufacture centrally, and distribute models. Cryobioprinted tissues address these challenges by enabling long-term storage and on-demand use. This advancement improves accessibility, reproducibility, and scalability for applications such as drug screening, toxicity testing, and disease modeling. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on cryobioprinting, a method that combines 3D bioprinting with integrated cryopreservation during biofabrication. The resulting constructs can be frozen, stored, shipped, and thawed while maintaining structural integrity and biological function. The platform supports diverse cell types and applications, including personalized medicine. In addition to supporting drug development workflows, the technology improves efficiency, accelerates timelines, and aims to enhance predictive accuracy while reducing overall drug development costs. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-09

Freely accessible, curated, and annotated clinical and physiological data – and tools to manage and analyze them – are foundational to fully harnessing the promise that machine learning/artificial intelligence (ML/AI) and signal and image processing hold for advancing and personalizing healthcare. Since 1999, PhysioNet has pioneered the free access to large, high-impact, de-identified physiological and clinical databases and associated analysis tools. PhysioNet is a recommended data repository for numerous publishers and the NIH, and its current content of over 350 diverse databases, software packages, and ML/AI models is used extensively across the world. In the past decade, the PhysioNet user community has grown dramatically: In 2023 alone, ~30,000 new users registered for accounts, nearly 7,000 publications referenced PhysioNet and its contents, and over 80 new databases or software packages were published on the platform. With such rapid growth, PhysioNet faces significant challenges in meeting the evolving needs of the community and in managing the growing number of new submissions. Balancing these demands while upholding a rigorous editorial review process is essential to ensure that published projects maintain the high standards of quality expected by the research community. To be able to review, curate, and publish the growing number of contributions submitted to PhysioNet in a timely manner, and to meet the dramatically increasing demand for high-quality databases and analysis tools, the PhysioNet team must adapt its operations. With funding through this U24 mechanism, the team proposes to 1) Streamline access to high-quality content by decreasing the time from contribution through curation to publication, enabling more efficient data discovery, and reducing the time for users to gain access to data resources; 2) Develop pathways for long-term access and sustainability by consolidating PhysioNet with MIT infrastructure and developing models for institutional memberships; and 3) Foster an inclusive and dynamic community that encourages active participation and collaboration among users, researchers, and developers by strengthening community engagement, by developing methods for impact assessment, and by aligning resources and resource development with user needs. Attaining these aims will put PhysioNet on a sustained path of growth, enabling it to keep pace with the exponentially increasing demand for its resources.

NSF Awards · FY 2025 · 2025-09

Nontechnical description: Bismuth is both an element and a material that has attracted tremendous interest from scientific researchers throughout history. For example, it was the first metal whose Fermi surface was experimentally identified and its study has led to the discovery of quantum (e.g. Shubnikov-de Haas) oscillations. It is also deemed a ‘magic element’ by chemists as it has rare chemical properties allowing it to form compounds with a diverse range of nuclei. A few years ago, the PI’s group made the discovery that ultralow contact resistance can be made to transition metal dichalcogenide (TMD, e.g. MoS2) devices when Bi is deposited on them. Initial structural analysis revealed that Bi formed epitaxial structures on TMD, which was understood as a particular semi-metallic phase of Bi at the time the work was published. Nevertheless, further recent studies indicated that previous understanding might be mistaken, which motivated this research to solve key mysteries and advance scientific knowledge. The findings will lead to discoveries of new 2D forms of materials and enable low power, high-performance devices for both conventional and quantum computing. The project will provide lab experience to undergraduate students and outreach to high school and other students. Technical description: This project will systematically investigate the phase and structures that Bi forms at the 2D material interface under various controlled conditions and will characterize the electrical and optical properties of the resulting structures. A layer-by-layer characterization will be carried out to examine the interfacial properties for one layer and multiple layers, to identify if any transitions occur as layer thickness increases. The unexpected and unexplored phenomena of Bi phase formation on a 2D substrate have revealed a knowledge gap in the 2D research field. Together with the particularly interesting outcomes – low contact resistance, much better thermal & chemical stability, and unusually high doping – these studies will not only provide a deeper understanding of the interfacial phenomena, but will also inspire other investigations of materials formation on 2D templates, and may have significant impact into technologically important areas such as electrical contact formation on 2D devices, high temperature quantum spin Hall materials, etc. The study of novel material phases and their formation will open new research directions, enabling new technologies for a wide range of applications, including energy, catalysis and biomedical. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

PRIMES (Program for Research in Mathematics, Engineering and Science) will select talented high school students from across the U.S. via rigorous testing. Under the guidance of academic mentors, program participants will conduct year-long research projects, write papers, and make conference presentations. Groups of students will also participate in guided reading, online research forums, and a residential summer math camp. The program will create a pipeline of mathematical talent and support graduate students and undergraduates serving as mentors. Topics for student research projects will include ancient ALE Ricci flows and dynamical energy functionals, refractive outer billiards, fields of definition of abelian surfaces of maximal Picard rank, semisimplifications of representations of gl(n), Temperley-Lieb algebras and canonical bases, tournament and digraph inversions, machine learning for physical systems, sparse inference of earthquake dynamics, and Fresnel inversion and the NASA Cassini Mission. More details and information about the program may be found on the PRIMES website: https://math.mit.edu/research/highschool/primes. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

The lifecycle of many marine organisms includes a free-swimming larval stage that searches for a place to attach, where it metamorphoses into an adult. This search and decision-making process (larval settlement) is poorly understood yet essential: it is expected to play critical roles in determining species density and to be point of vulnerability. Understanding larval settlement therefore has important societal implications for both promoting the growth of desired species and protecting regions from invasive species and biofouling. Despite the importance of this process, remarkably little is known of the underlying neural mechanisms. Using larvae from the jellyfish Clytia as a model, this work will perform a systematic study of the neural control of settlement. This deeper understanding could be foundational for developing new solutions for prediction and manipulation of the density of critical marine organisms for the betterment of human society and ocean health. In addition to these scientific impacts, integral to this proposed work is training the next generation of scientists. Clytia make an exceptional research organism for undergraduate research: it is easy and low-cost for students to gain hands-on research experience using Clytia, and this work will expose numerous undergraduates to research experience. Lastly, a goal of this work is for Clytia to become a widely used model organism for neuroscience research. This work will be foundational for bolstering a Clytia community asking fundamental questions in neural evolution, development, regeneration, and the neural control of behavior. The tools and approaches developed here are also expected to be readily applied to other species. Together, the impacts of this work are expected to go beyond the scientific research to include education, outreach, bolstering a broader scientific community, and other benefits to society. The technical approach to study the neural control of settlement proposed here is to leverage modern optical and genetic tools to ask: what is the logic that larvae use to explore their environments and how is this logic implemented in neural systems? Genetic tools have been established in Clytia, and this proposal applies a suite of modern neuroscience tools for behavioral analysis, imaging neural activity, and ablating specific neural cell types. In Aim 1, this proposal will apply automated behavioral tracking and modelling to delineate the strategy employed by larvae as they search for and choose a location to settle; it will then ask how these behaviors are impacted by environmental stressors. In Aim 2, this proposal will record from the entire larval nervous system to ask how sensory processing and decision-making are implemented in population neural activity. It will then examine which aspects of neural processing are vulnerable/resilient to perturbation. Lastly, in Aim 3, this proposal will examine the causal contributions of neural subnetworks through loss-of-function experiments. Together, this work is expected to contribute foundational insights into the neural control of behavior in these critically important marine organisms and to reveal aspects of their neurobiology that confer vulnerability or resistance in response to perturbation. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.