arXiv subjects

Christopher M. Rouleau

Publications and source records attributed to Christopher M. Rouleau.

9 recordsLinked to original sources

arXiv2026

An Open-Source Hardware and Software Toolkit to Enable Agentic RHEED-Guided Thin-Film Synthesis

Reflection high-energy electron diffraction (RHEED) provides rich information about evolving surfaces during thin-film growth, but non-automated, operator-dependent alignment and fragmented analysis workflows limit its potential in fully autonomous synthesis. Here, we present an open-source hardware and software toolkit that makes RHEED control and quantitative analysis accessible to operators and artificial intelligence (AI) agents. Demonstrated on a pulsed laser deposition system, the toolkit provides programmable electron-optics control, automated beam alignment and rocking-curve acquisition, and a training-free method for crystallographic azimuthal alignment. The auto RHEED application extracts structural and growth-related observables through shared graphical and programmatic interfaces, including a Model Context Protocol (MCP) server. An agent-driven demonstration shows how natural-language requests can guide the selection, configuration, and execution of quantitative analyses. An extensible adapter interface streamlines the incorporation of community-developed methods for AI analysis and RHEED simulation as they emerge, supported by Markdown implementation guides designed for AI coding agents. These capabilities support complementary descriptions of surface evolution through physical measurements and learned image representations while providing a practical route for incorporating new computational methods. Together, these tools reduce barriers to automated and agentic RHEED measurements and establish a foundation for future agentic control of thin-film synthesis guided by the evolving surface.

cond-mat.mtrl-sci↗

arXiv2026

Direct observation of ultrafast defect-bound and free exciton dynamics in defect-engineered WS$_2$ monolayers

Defects in two-dimensional transition metal dichalcogenides (TMDCs) broadly affect their optical and electronic properties. Directly capturing the ultrafast processes of exciton trapping and defect-bound exciton formation is crucial for understanding and advancing defect-mediated optoelectronics and quantum technologies. However, the weak transient optical absorption of defect-bound excitons has limited their experimental observation to date. Here, we report the direct observation of the ultrafast dynamics of defect-bound excitons in monolayer WS$_2$ crystals with a high density of mono-sulfur vacancies (V$_S$) and W-site defect complexes (S$_W$V$_S$) resulting from synthesis by alkali metal halide-assisted chemical vapor deposition. The dynamics of excitons bound to these defects, along with their coherent interactions with free excitons, are elucidated using ultrafast optical spectroscopy. Using above band-edge photoexcitation, we find that both free and defect-bound excitons simultaneously form within 300 fs from hot carrier relaxation. The defect-bound excitons exhibit shorter lifetimes than free excitons, leading to a population difference of the corresponding excitonic states and free exciton trapping within a 1--100 ps window. Band-edge photoexcitation of free and defect-bound exciton states reveals ultrafast interconversion within ~150 fs (comparable to our temporal resolution), indicating possible coherent coupling between these states. We further demonstrate efficient up-conversion of defect-bound excitons to free excitons with photon energies up to ~300 meV below the free exciton resonance. These findings provide insights into the ultrafast dynamics of defect-bound excitons in TMDCs and their coupling with free excitons, which are relevant to defect-engineered optoelectronic, quantum photonic, and valleytronic applications.

physics.optics↗

arXiv2026

Spontaneous Polarization Suppression of Exciton-Exciton Annihilation in 3R-Stacked MoS$_2$ Bilayers

Rapid exciton-exciton annihilation (EEA) in two-dimensional semiconductors limits access to high-density excitonic regimes essential for efficient optoelectronic operation under strong excitation. Here, we show that EEA is suppressed by repulsive dipole-dipole interactions between interlayer excitons polarized by the spontaneous polarization intrinsic to rhombohedral (3R)-stacked MoS$_2$ bilayers. Using ultrafast pump-probe spectroscopy, we measure an EEA rate of $γ_{\rm EEA}=(5.03\pm0.99)\times10^{-3}$ cm$^2$ s$^{-1}$ in 3R bilayers, which is approximately 18.2-fold smaller than that in monolayers and 2.9-fold smaller than that in nonpolar 2H bilayers. Despite the higher exciton diffusivity recently reported for 3R relative to 2H bilayers, the reduced EEA rate in 3R indicates a rate-limited regime governed by the close-encounter annihilation probability rather than diffusion. A rate-limited annihilation model incorporating a dipole-dipole repulsive potential captures the observed ratio $γ_{{\rm EEA},3{\rm R}}/γ_{{\rm EEA},2{\rm H}}\approx0.35$ for an exciton-exciton encounter distance of $\sim$1.3 nm, consistent with the bilayer exciton Bohr radius. These results show that spontaneous polarization in 3R-stacked bilayers suppresses nonlinear excitonic losses and provides a route toward high-density excitonics.

cond-mat.mes-hall↗

arXiv2025

Human-AI collaborative autonomous synthesis with pulsed laser deposition for remote epitaxy

Autonomous laboratories typically rely on data-driven decision-making, occasionally with human-in-the-loop oversight to inject domain expertise. Fully leveraging AI agents, however, requires tightly coupled, collaborative workflows spanning hypothesis generation, experimental planning, execution, and interpretation. To address this, we develop and deploy a human-AI collaborative (HAIC) workflow that integrates large language models for hypothesis generation and analysis, with collaborative policy updates driving autonomous pulsed laser deposition (PLD) experiments for remote epitaxy of BaTiO$_3$/graphene. HAIC accelerated the hypothesis formation and experimental design and efficiently mapped the growth space to graphene-damage. In situ Raman spectroscopy reveals that chemistry drives degradation while the highest energy plume components seed defects, identifying a low-O$_2$ pressure low-temperature synthesis window that preserves graphene but is incompatible with optimal BaTiO$_3$ growth. Thus, we show a two-step Ar/O$_2$ deposition is required to exfoliate ferroelectric BaTiO$_3$ while maintaining a monolayer graphene interlayer. HAIC stages human insight with AI reasoning between autonomous batches to drive rapid scientific progress, providing an evolution to many existing human-in-the-loop autonomous workflows.

cond-mat.mtrl-sci↗

arXiv2025

Deep learning with reflection high-energy electron diffraction images to predict cation ratio in Sr$_{2x}$Ti$_{2(1-x)}$O$_{3}$ thin films

Machine learning (ML) with in situ diagnostics offers a transformative approach to accelerate, understand, and control thin film synthesis by uncovering relationships between synthesis conditions and material properties. In this study, we demonstrate the application of deep learning to predict the stoichiometry of Sr$_{2x}$Ti$_{2(1-x)}$O$_{3}$ thin films using reflection high-energy electron diffraction images acquired during pulsed laser deposition. A gated convolutional neural network trained for regression of the Sr atomic fraction achieved accurate predictions with a small dataset of 31 samples. Explainable AI techniques revealed a previously unknown correlation between diffraction streak features and cation stoichiometry in Sr$_{2x}$Ti$_{2(1-x)}$O$_{3}$ thin films. Our results demonstrate how ML can be used to transform a ubiquitous in situ diagnostic tool, that is usually limited to qualitative assessments, into a quantitative surrogate measurement of continuously valued thin film properties. Such methods are critically needed to enable real-time control, autonomous workflows, and accelerate traditional synthesis approaches.

cond-mat.mtrl-sci↗

arXiv2023

Deep learning with plasma plume image sequences for anomaly detection and prediction of growth kinetics during pulsed laser deposition

Materials synthesis platforms that are designed for autonomous experimentation are capable of collecting multimodal diagnostic data that can be utilized for feedback to optimize material properties. Pulsed laser deposition (PLD) is emerging as a viable autonomous synthesis tool, and so the need arises to develop machine learning (ML) techniques that are capable of extracting information from in situ diagnostics. Here, we demonstrate that intensified-CCD image sequences of the plasma plume generated during PLD can be used for anomaly detection and the prediction of thin film growth kinetics. We developed a multi-output (2$+$1)D convolutional neural network regression model that extracts deep features from plume dynamics that not only correlate with the measured chamber pressure and incident laser energy, but more importantly, predict parameters of an auto-catalytic film growth model derived from in situ laser reflectivity experiments. Our results are the first demonstration of how ML with in situ plume diagnostics data in PLD can be utilized to maintain deposition conditions in an optimal regime. Further, the predictive capabilities of plume dynamics on the kinetics of film growth or other film properties prior to deposition provides a means for rapid pre-screening of growth conditions for the non-expert, which promises to accelerate materials optimization with PLD.

cond-mat.mtrl-sci↗

arXiv2023

Autonomous synthesis of thin film materials with pulsed laser deposition enabled by in situ spectroscopy and automation

Synthesis of thin films has traditionally relied upon slow, sequential processes carried out with substantial human intervention, frequently utilizing a mix of experience and serendipity to optimize material structure and properties. With recent advances in autonomous systems which combine synthesis, characterization, and decision making with artificial intelligence (AI), large parameter spaces can be explored autonomously at rates beyond what is possible by human experimentalists, greatly accelerating discovery, optimization, and understanding in materials synthesis which directly address the grand challenges in synthesis science. Here, we demonstrate autonomous synthesis of a contemporary 2D material by combining the highly versatile pulsed laser deposition (PLD) technique with automation and machine learning (ML). We incorporated in situ and real-time spectroscopy, a high-throughput methodology, and cloud connectivity to enable autonomous synthesis workflows with PLD. Ultrathin WSe2 films were grown using co-ablation of two targets and showed a 10x increase in throughput over traditional PLD workflows. Gaussian process regression and Bayesian optimization were used with in situ Raman spectroscopy to autonomously discover two distinct growth windows and the process-property relationship after sampling only 0.25% of a large 4D parameter space. Any material that can be grown with PLD could be autonomously synthesized with our platform and workflows, enabling accelerated discovery and optimization of a vast number of materials.

cond-mat.mtrl-sci↗

arXiv2016

Growth control of oxygen stoichiometry in homoepitaxial SrTiO3 films by pulsed laser epitaxy in high vacuum

In many transition metal oxides (TMOs), oxygen stoichiometry is one of the most critical parameters that plays a key role in determining the structural, physical, optical, and electrochemical properties of the material. However, controlling the growth to obtain high quality single crystal films having the right oxygen stoichiometry, especially in a high vacuum environment, has been viewed as a challenge. In this work, we show that through proper control of the plume kinetic energy, stoichiometric crystalline films can be synthesized without generating oxygen defects, even in high vacuum. We use a model homoepitaxial system of SrTiO3 (STO) thin films on single crystal STO substrates. Physical property measurements indicate that oxygen vacancy generation in high vacuum is strongly influenced by the energetics of the laser plume, and it can be controlled by proper laser beam delivery. Therefore, our finding not only provides essential insight into oxygen stoichiometry control in high vacuum for understanding the fundamental properties of STO-based thin films and heterostructures, but expands the utility of pulsed laser epitaxy of other materials as well.

cond-mat.mtrl-sci↗

arXiv2012

Atomic layer engineering of perovskite oxides for chemically sharp heterointerfaces

Advances in synthesis techniques and materials understanding have given rise to oxide heterostructures with intriguing physical phenomena that cannot be found in their constituents. In these structures, precise control of interface quality, including oxygen stoichiometry, is critical for unambiguous tailoring of the interfacial properties, with deposition of the first monolayer being the most important step in shaping a well-defined functional interface. Here, we studied interface formation and strain evolution during the initial growth of LaAlO3 on SrTiO3 by pulsed laser deposition, in search of a means for controlling the atomic-sharpness of the interfaces. Our experimental results show that growth of LaAlO3 at a high oxygen pressure dramatically enhances interface abruptness. As a consequence, the critical thickness for strain relaxation was increased, facilitating coherent epitaxy of perovskite oxides. This provides a clear understanding of the role of oxygen pressure during the interface formation, and enables the synthesis of oxide heterostructures with chemically-sharper interfaces.

cond-mat.mtrl-sci↗