arXiv subjects

Sumner B. Harris

Publications and source records attributed to Sumner B. Harris.

18 recordsLinked to original sources

arXiv2026

AEcroscopyWave: Towards Self-Driving Characterization Platforms for Agentic AI

The characterization of electronic materials has traditionally been stratified into two distinct regimens: industry-scale automated systems to inspect materials for defects and ensure quality (such as in the semiconductor industry), and highly customized, operator-driven systems requiring human experts. The former offers high throughput but limited flexibility, whereas the latter is heavily bandwidth-limited but provides research-grade discovery capabilities. Recent advances in "self-driving" characterization tools offer the potential to bridge the two stratified regimes, by the creation of application program interfaces (APIs) that can control hardware, and the integration of AI methods to incorporate autonomy into the process. Here, we discuss our latest developments in AEcroscopyWave, a custom-built characterization platform for the agentic-AI era, that provides unified control of scanning probe microscopes with programmable peripheral instrumentation, highlighting the design choices that are necessary for maximizing the capability of the system and the ease of use for both human and AI agents. The benefits of making heterogeneous scientific instruments accessible, composable and usable by agents is demonstrated by test cases.

cond-mat.mtrl-sci↗

arXiv2026

Active Learning for Tractable and Reproducible Pulsed Laser Deposition

This paper shows how data-driven machine learning approaches can improve growth control, reproducibility, and physical insight in the pulsed laser deposition (PLD) growth of correlated oxides. Despite well-known relationships between growth conditions and material properties, consistently producing high-quality films of complex materials like LaVO$_3$ remains difficult due to the highly non-equilibrium nature of PLD and the defects and competing phases that accumulate during growth. Here, we use an active learning framework based on Gaussian process Bayesian optimization that incorporates measured bulk and surface lattice properties along with impurity phase information to efficiently map the multidimensional growth space of LaVO$_3$ by PLD. By tuning the relative weighting of these properties, the model identifies an optimized region where phase-pure films of LaVO$_3$ exhibit two-dimensional surfaces, near-ideal lattice parameters, and minimal sub-band gap optical absorption. The trained model reveals clear competition among different defect formation mechanisms that are connected to unseen parameters like supersaturation and surface mobility, thus giving insight into the highly non-equilibrium process of PLD growth. Together, this demonstrates that property-guided machine learning can accelerate materials optimization while providing a new way to address fundamental growth mechanisms in PLD that enable understanding and utilization of quantum phenomena found in complex oxides.

cond-mat.mtrl-sci↗

arXiv2025

Thermal and Size Effects in Ferroelastic Domains by Machine Learning

Ferroelastic domain walls (DWs) underpin key functionalities in complex oxides. In free-standing ferroic thin films, where elastic interactions are highly thickness dependent, understanding DW behaviour across length scales and external stimuli is crucial. A thickness-dependent monopolar-to-dipolar crossover in elastic DW behaviour has been reported; however, how temperature influences this regime remains unexplored. Here, LaAlO3 thin films spanning the dipolar ($<200$ nm) and crossover (200-300 nm) regimes are investigated using in situ heating scanning transmission electron microscopy (STEM) and a machine-learning-driven image analysis approach. By tracking DW curvature and density from above $T_C$ (approximately $550,^\circ$C) to room temperature (RT), a distinct interplay between temperature and thickness is identified. In the dipolar regime, DWs are mobile and curved near $T_C$ and gradually freeze upon cooling, consistent with the well-known temperature freezing regime. In contrast, within the crossover regime, DWs are nearly static, with minimal reconfiguration through cooling and curvature an order of magnitude lower at RT. These results map the evolution of DWs across the thermally driven super-elastic to freezing regimes, revealing how thickness and temperature govern DW morphology and dynamics, and providing insight relevant for domain engineering in free-standing oxide thin films.

cond-mat.mtrl-sci↗

arXiv2025

Auto-3DPFM: Automating Polarization-Vector Mapping at the Nanoscale

The functional properties of ferroelectric materials are strongly influenced by ferroelectric polarization orientation; as such, access to consistent and precise characterization of polarization vectors is of substantial importance to ferroelectrics research. Here, we develop a fully automated three-dimensional piezoresponse force microscopy (Auto-3DPFM) technique automating all essential steps in interferometric PFM for 3D polarization vector characterization, including laser alignment, tip calibration and approach, image acquisition, polarization vector reconstruction, and visualization. The automation reduces the experimental burden of ferroelectric polarization vector characterization, while the back-and-forth calibration ensures consistency and reproducibility of 3D polarization reconstruction. An algorithmic workflow is also developed to identify domain walls and calculate their characteristic angles via a spatial vector-angle-difference method, presenting one unique capability enabled by Auto-3DPFM that is not accessible with traditional PFM techniques. Beyond representing a significant step forward in 3D polarization mapping, Auto-3DPFM promises to accelerate discovery via high-throughput and autonomous characterization in ferroelectric materials research. When integrated with machine learning and adaptive sampling strategies in self-driving labs, Auto-3DPFM will serve as a valuable tool for advancing ferroelectric physics and microelectronics development.

cond-mat.mtrl-sci↗

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

Pulsed laser synthesis of mesoporous metal chalcogenide thin films

Mesoporous films of the metal chalcogenide $\beta$-FeSe were grown on MgO substrates by KrF pulsed laser deposition (PLD) in an argon background. At 100 mTorr, gated intensified charge-coupled device imaging and ion probe measurements showed that the plasma plume responsible for crystal growth initially comprised three components, with distinct expansion velocities. Plume interactions with the substrate heater and ablation target gave rise to complex dynamics, including collisions between the charged leading edge -- rebounding between the substrate and the target -- and slower-moving species in the plume interior. Film growth was dominated by species with kinetic energies $\le$0.5 eV/atom. X-ray reflectivity and atomic force microscopy revealed that films grown in this environment -- with a substrate temperature of 350$^\circ$C, a laser fluence of 1.0 J cm$^{-2}$, and a 7.5 mm$^2$ spot area -- formed a porous framework with 15% porosity and pore sizes below 100 nm. X-ray diffraction indicated that the porous films were epitaxial with respect to the substrate and likely grew by oriented-attachment of gas-phase molecular clusters or very small nanoparticles, in contrast to the conventional epitaxy of vacuum films from atomic constituents. The in-plane orientation of the mesoporous films was $\beta$-FeSe[100]$\parallel$[110]MgO, attributed to the soft landing of pre-formed crystallites on the MgO substrates, where protruding Se rows of $\beta$-FeSe aligned with corrugations of the MgO surface. This work implies that growth of candidate electrocatalyst materials by PLD in inert gas background may allow mesoporous frameworks with a single crystallographic orientation that expose specific crystal facets for electrochemical reactions and active site engineering.

cond-mat.mtrl-sci↗

arXiv2025

Ferroelastic Domain Induced Electronic Modulation in Halide Perovskites

Lead halide perovskites have emerged as promising materials for optoelectronic applications due to their exceptional properties. In the all-inorganic CsPbBr3 perovskites, ferroelastic domains formed during phase transitions enhance bulk transport and emissive efficiency. However, the microscopic mechanisms governing carrier dynamics remain poorly understood. In this study, we employ cathodoluminescence (CL) and micro-Raman spectroscopy to image and investigate the electronic properties of the ferroelastic domain walls in CsPbBr3 single crystals. CL measurements reveal a reduced emissive yield and a slight redshift in emission at the domain walls. Further, micro-Raman studies provide spatially resolved mapping of vibrational modes, exhibiting second-order phonon modes localized at the domain boundaries. Our findings suggest that electron-phonon coupling at twin domain walls plays a critical role in facilitating efficient charge separation, thereby improving the optoelectronic performance of the CsPbBr3 perovskites.

cond-mat.mtrl-sci↗

arXiv2025

Atom identification in bilayer moire materials with Gomb-Net

Moire patterns in van der Waals bilayer materials complicate the analysis of atomic-resolution images, hindering the atomic-scale insight typically attainable with scanning transmission electron microscopy. Here, we report a method to detect the positions and identities of atoms in each of the individual layers that compose twisted bilayer heterostructures. We developed a deep learning model, Gomb-Net, which identifies the coordinates and atomic species in each layer, effectively deconvoluting the moire pattern. This enables layer-specific mapping of quantities like strain and dopant distributions, unlike other commonly used segmentation models which struggle with moire-induced complexity. Using this approach, we explored the Se atom substitutional site distribution in a twisted fractional Janus WS2-WS2(1-x)Se2x heterostructure and found that layer-specific implantation sites are unaffected by the moire pattern's local energetic or electronic modulation. This advancement enables atom identification within material regimes where it was not possible before, opening new insights into previously inaccessible material physics.

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↗

arXiv2024

Bayesian state estimation unlocks real-time control in thin film synthesis

The rapid validation of newly predicted materials through autonomous synthesis requires real-time adaptive control methods that exploit physics knowledge, a capability that is lacking in most systems. Here, we demonstrate an approach to enable the real-time control of thin film synthesis by combining in situ optical diagnostics with a Bayesian state estimation method. We developed a physical model for film growth and applied the Direct Filter (DF) method for real-time estimation of nucleation and growth rates during pulsed laser deposition (PLD) of transition metal dichalcogenides. We validated the approach on simulated and previously acquired reflectivity data for WSe$_2$ growth and ultimately deployed the algorithm on an autonomous PLD system during growth of 1T$^\prime$-MoTe$_2$ under various synthesis conditions. We found that the DF robustly estimates growth parameters in real-time at early stages of growth, down to 15% percent monolayer area coverage. This approach opens new opportunities for adaptive film growth control based on a fusion of in situ diagnostics, modern data assimilation methods, and physical models which promises to enable control of synthesis trajectories towards desired material states.

cond-mat.mtrl-sci↗

arXiv2024

Active oversight and quality control in standard Bayesian optimization for autonomous experiments

The fusion of experimental automation and machine learning has catalyzed a new era in materials research, prominently featuring Gaussian Process Bayesian Optimization (GPBO) driven autonomous experiments navigating complex experimental conditions for accelerated scientific discovery. In traditional GPBO-driven experiments, a predefined scalarizer function is often required to preprocess the experimental data, transforming non-scalar raw data into scalar descriptors for GP training. However, such predefined scalarizer functions have limitations, which likely fail to accommodate the diversity and complexity of real-world experimental data, potentially skewing experimental outcomes. Thus, oversight and quality control are necessitated over the process to avoid GPBO from being misled by low quality scalarizers. To address the limitation, we introduce a Dual-GP approach that enhances traditional GPBO by adding a secondary surrogate model to dynamically constrain the experimental space based on real-time assessments of the raw experimental data. This Dual-GP approach enhances the optimization efficiency of traditional GPBO by isolating more promising space for BO sampling and more valuable experimental data for primary GP training. We also incorporate a flexible, human-in-the-loop intervention method in the Dual-GP workflow to adjust for unanticipated results. We demonstrate the effectiveness of the Dual-GP model with synthetic model data and implement this approach in autonomous pulsed laser deposition experimental data. This Dual-GP approach has broad applicability in diverse GPBO-driven experimental settings, providing a more adaptable and precise framework for refining autonomous experimentation for more efficient optimization.

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↗

arXiv2023

Machine learning the relationship between Debye temperature and superconducting transition temperature

Recently a relationship between the Debye temperature $\Theta_D$ and the superconducting transition temperature $T_c$ of conventional superconductors has been proposed [npj Quantum Materials $\mathbf{3}$, 59 (2018)]. The relationship indicates that $T_c \le A \Theta_D$ for phonon-mediated BCS superconductors, with $A$ being a pre-factor of order $\sim 0.1$. In order to verify this bound, we train machine learning (ML) models with 10,330 samples in the Materials Project database to predict $\Theta_D$. By applying our ML models to 9,860 known superconductors in the NIMS SuperCon database, we find that the conventional superconductors in the database indeed follow the proposed bound. We also perform first-principles phonon calculations for H$_{3}$S and LaH$_{10}$ at 200 GPa. The calculation results indicate that these high-pressure hydrides essentially saturate the bound of $T_c$ versus $\Theta_D$.

cond-mat.supr-con↗

arXiv2021

Geometrical and energy scaling in the pulsed laser deposition plasma during epitaxial growth of FeSe thin films

Pulsed laser deposition (PLD) is a versatile technique for growing epitaxial heterostructures of a wide variety of novel materials combinations. Achieving low-defect layers with atomically sharp interfaces by PLD requires careful management of crystal growth conditions. Control over the flux of depositing particles, their kinetic energy, and the substrate temperature, is generally sufficient to obtain high-quality single crystal epitaxy in vacuum. In this article, we show that measurements of plasma parameters such as the electron temperature, the electron density, and the Mach number of the plasma fluid expansion may provide additional insights into the tuning of the crystal growth and processing environment of PLD. We report Langmuir probe measurements during growth of FeSe thin films on (100)-oriented SrTiO$_3$. We discuss two distinct plasma regimes that are accessible by KrF laser ablation of FeSe when pulse energies in the 20-630 mJ range are used. The two regimes can be characterized by the relative number of photons to absorbing centers in the laser-plasma interaction volume. Thin films grown under the conditions created by these two distinct plasma regimes are analyzed by x-ray diffraction and x-ray reflectivity and their epitaxial configurations correlated to the PLD conditions associated with these plasmas.

physics.app-ph↗

arXiv2021

Experimentally constrained multidimensional simulation of laser-generated plasmas and its application to UV nanosecond ablation of Se and Te

We carry out simulations of laser plasmas generated during UV nanosecond pulsed laser ablation of the chalcogens selenium (Se) and tellurium (Te), and compare the results to experiments. We take advantage of a 2D-axisymmetric, adaptive Cartesian Mesh (ACM) framework that enables plume simulations out to centimeter distances over tens of microseconds. Our model and computational technique enable comparison to laser-plasma applications where the long-term behavior of the plume is of primary interest, such as pulsed laser synthesis and modification of materials. An effective plasma absorption term is introduced in the model, allowing the simulation to be constrained by experimental time-of-flight kinetic energy distributions. We show that the effective simulation qualitatively captures the key characteristics of the observed laser plasma, including the effect of laser spot size. Predictions of full-scale experimentally-constrained Se and Te plasmas for 4.0 J/cm$^2$ laser fluence and 1.8 mm$^2$ circular laser spot area show distinct behavior compared to more commonly studied copper (Cu) plumes. The chalcogen plumes have spatial gradients of plasma density that are steeper than those for Cu by up to three orders of magnitude. Their spatial ion distributions have central bulges, in contrast to the edge-only ionization of Cu. For the irradiation conditions explored, the range of plasma temperatures for Se and Te is predicted to be higher than for Cu by more than 0.50 eV.

physics.plasm-ph↗

arXiv2019

Laser-generated plasmas in length scales relevant for thin film growth and processing: simulation and experiment

In pulsed laser deposition, thin film growth is mediated by a laser-generated plasma, whose properties are critical for controlling the film microstructure. The advent of 2D materials has renewed the interest in how this ablation plasma can be used to manipulate the growth and processing of atomically thin systems. For such purpose, a quantitative understanding of the density, charge state, and kinetic energy of plasma constituents is needed at the location where they contribute to materials processes. Here we study laser-induced plasmas over expansion distances of several centimeters from the ablation target, which is the relevant length scale for materials growth and modification. The study is enabled by a fast implementation of a laser ablation/plasma expansion model using an adaptive Cartesian mesh solver. Simulation outcomes for KrF excimer laser ablation of Cu are compared with Langmuir probe and optical emission spectroscopy measurements. Simulation predictions for the plasma-shielding threshold, the ionization state of species in the plasma, and the kinetic energy of ions, are in good correspondence with experimental data. For laser fluences of 1-4 J/cm$^2$, the plume is dominated by Cu$^0$, with small concentrations of Cu$^{+}$ and electrons at the expansion front. Higher laser fluences (e.g., 7 J/cm$^2$) lead to a Cu$^{+}$-rich plasma, with a fully ionized leading edge where Cu$^{2+}$ is the dominant species. In both regimes, simulations indicate the presence of a low-density, high-temperature plasma expansion front with a high degree of ionization that may play a significant role in doping, annealing, and kinetically-driven phase transformations in 2D materials.

physics.app-ph↗

arXiv2019

X-ray diffraction analysis to support phase identification in FeSe and Fe$_7$Se$_8$ epitaxial thin films

X-ray diffraction (XRD) data and analysis for epitaxial iron selenide thin films grown by pulsed laser deposition (PLD) are presented. The films contain $β$-FeSe and Fe$_7$Se$_8$ phases in a double epitaxy configuration with the $β$-FeSe phase (001) oriented on the (001) MgO growth substrate. Fe$_7$Se$_8$ simultaneously takes on two different epitaxial orientations in certain growth conditions, exhibiting both (101)- and (001)- orientations. Each of these orientations are verified with the presented XRD data. Additionally, XRD data used to determine the PLD target composition as well as mosaic structure of the $β$-FeSe phase are shown.

cond-mat.mtrl-sci↗