Collaborative research: dmref: NSF-DFG: AI-driven platform for 2D materials synthesis and discovery

The National Science Foundation Division of Materials Research awarded Kenyon College $160,000 on October 1, 2026, under the Mathematical and Physical Sciences program (CFDA 47.049) to develop an AI-guided platform for two-dimensional materials synthesis and discovery.

The research integrates real-time optical diagnostics, materials characterization, multiscale simulation, and machine learning to establish predictive strategies for controlling two-dimensional semiconductor synthesis via metalorganic chemical vapor deposition. A core innovation is in-situ spectroscopic ellipsometry that continuously monitors atomically thin film evolution during growth, providing immediate feedback to computational models to rapidly identify promising synthesis pathways. The project combines advanced thin-film synthesis, real-time monitoring, characterization, multiscale simulation spanning density functional theory and reactive molecular dynamics, and data science to uncover fundamental relationships governing two-dimensional materials synthesis. The work includes interdisciplinary training at the interface of materials science, data science, and artificial intelligence, with open sharing of data, software, and educational resources to the broader research community.

Period of performance runs from October 1, 2026, through September 30, 2030. Place of performance is Gambier, Ohio. This is a collaborative research award under the Designing Materials to Revolutionize and Engineer our Future (DMREF) program in partnership with the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG).

Abstract

Non-technical description two-dimensional (2D) semiconductors are promising materials for future electronic technologies because they can enable smaller, faster, and more energy-efficient devices than conventional silicon. However, producing these atomically thin materials reproducibly over large wafer areas remains a major scientific and manufacturing challenge. This project develops new artificial intelligence (AI)-guided approaches that combine advanced thin-film synthesis, real-time monitoring of film growth, characterization, multiscale simulation, and data science to uncover the fundamental relationships governing two-dimensional materials synthesis and establishes predictive strategies for controlling film growth and materials properties. A key innovation is the development of a real-time optical diagnostic that continuously monitors how the atomically thin film evolves during growth, providing immediate feedback that can be combined with computational models to rapidly identify promising synthesis pathways. By integrating real-time measurements with predictive simulations and machine learning, the research aims to accelerate the discovery of high-quality 2D semiconductors while providing new scientific insights into the complex processes that govern their formation. The project also provides interdisciplinary training at the interface of materials science, data science, and artificial intelligence while openly sharing data, software, and educational resources to benefit the broader research community. Technical description the research establishes an AI-guided framework for metalorganic chemical vapor deposition (MOCVD) of wafer-scale transition metal dichalcogenide semiconductors. The project integrates in-situ spectroscopic ellipsometry, EX-situ materials characterization, multiscale simulations spanning density functional theory, reactive molecular dynamics, and computational fluid dynamics, and multi-task, multi-fidelity machine learning to uncover the fundamental relationships linking precursor chemistry, transport phenomena, nucleation, film growth, and resulting materials properties. Bayesian optimization combines experimental and simulation data to efficiently explore new precursor chemistries and identify synthesis pathways that maximize crystal quality, layer uniformity, and wafer-scale reproducibility. By connecting atomic-scale reaction mechanisms with reactor-scale synthesis and experimentally measured materials properties, the project establishes predictive synthesis-structure-property relationships and new design principles for scalable manufacturing of two-dimensional semiconductors. This integrated approach directly supports the goals of the designing materials to revolutionize and engineer our future (DMREF) program by accelerating materials discovery through the integration of experiment, computation, and data science. 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.- subawards are not planned for this award.