Autonomous systems that combine synthesis, characterization, and artificial intelligence can greatly accelerate the discovery and optimization of materials, however platforms for growth of macroscale thin films by physical vapor deposition techniques have lagged far behind others. Here this study demonstrates autonomous synthesis by pulsed laser deposition (PLD), a highly versatile synthesis technique, in the growth of ultrathin WSe₂ films. By combing the automation of PLD synthesis and in situ diagnostic feedback with a high‐throughput methodology, this study demonstrates a workflow and platform which uses Gaussian process regression and Bayesian optimization to autonomously identify growth regimes for WSe₂ films based on Raman spectral criteria by efficiently sampling 0.25% of the chosen 4D parameter space. With throughputs at least 10x faster than traditional PLD workflows, this platform and workflow enables the accelerated discovery and autonomous optimization of the vast number of materials that can be synthesized by PLD.

The fusion of experimental automation and machine learning has catalyzed a new era in materials research, prominently featuring Gaussian Process (GP) Bayesian Optimization (BO) driven autonomous experiments. Here 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.

Recently a relationship between the Debye temperature ${\mathrm{\Theta }}_D$ and the superconducting transition temperature $T_c$ of conventional superconductors has been proposed [Esterlis et al., npj Quantum Mater. 3, 59 (2018)]. The relationship indicates that $T_c\le A{\mathrm{\Theta }}_D$ for phonon-mediated BCS superconductors, with $A$ being a prefactor 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 ${\mathrm{\Theta }}_D$. By applying our ML models to 9860 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 ${\mathrm{H}}_3\mathrm{S}$ and ${\mathrm{LaH}}_{10}$ at 200 GPa. The calculation results indicate that these high-pressure hydrides essentially saturate the bound of $T_c$ versus ${\mathrm{\Theta }}_D$.

Isotope effects have received increasing attention in materials science and engineering because altering isotopes directly affects phonons, which can affect both thermal properties and optoelectronic properties of conventional semiconductors. However, how isotopic mass affects the optoelectronic properties in 2D semiconductors remains unclear because of measurement uncertainties resulting from sample heterogeneities. Here, we report an anomalous optical bandgap energy red shift of 13 (±7) milli–electron volts as mass of Mo isotopes is increased in laterally structured ¹⁰⁰ MoS ₂ - ⁹² MoS ₂ monolayers grown by a two-step chemical vapor deposition that mitigates the effects of heterogeneities. This trend, which is opposite to that observed in conventional semiconductors, is explained by many-body perturbation and time-dependent density functional theories that reveal unusually large exciton binding energy renormalizations exceeding the ground-state renormalization energy due to strong coupling between confined excitons and phonons. The isotope effect on the optical bandgap reported here provides perspective on the important role of exciton-phonon coupling in the physical properties of two-dimensional materials.

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 develop multi-output (2 + 1)D convolutional neural network regression models that extract 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 demonstrate 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.