SearcharxivSearch

arXiv subjects

Isaac Armstrong

Publications and source records attributed to Isaac Armstrong.

2 recordsLinked to original sources

Stereochemical Vacuum Gap Explains Out-of-Plane Thermal Insulation in MXenes

Two-dimensional MXenes are promising materials for thermal management and spectral camouflage, combining low out-of-plane thermal conductivity with low infrared emissivity and mechanical robustness. Yet the near-order-of-magnitude spread in experimental out-of-plane thermal conductivity measurements (0.14-0.8 W/mK) and the systematic overestimation by simulations point to a fundamental gap in our understanding of heat transport in these materials. Here, we argue these differences originate in the overlooked role of heterogeneous surface terminations. Using Non-Equilibrium Molecular Dynamics simulations of Ti3C2Tx, we show that this discrepancy arises from a stereochemically induced vacuum gap between adjacent layers, formed when surface terminations of different sizes coexist. Even minor deviations from homogeneous terminations drastically suppress out-of-plane thermal conductivity, bringing simulated values into quantitative agreement with experiment. We also show that thermal conductivity scales strongly with the atomic density, and that introducing bulky surface terminations, including residual water, reduces the thermal conductivity to 0.3 W/mK, an order of magnitude below homogeneous termination values and below the minimum thermal conductivity limit predicted for disordered solids. Thus, we propose a chemistry-driven route to engineer thermal transport in MXenes.

cond-mat.mtrl-sci

Enhancing Dimensionality Prediction in Hybrid Metal Halides via Feature Engineering and Class-Imbalance Mitigation

We present a machine learning framework for predicting the structural dimensionality of hybrid metal halides (HMHs), including organic-inorganic perovskites, using a combination of chemically-informed feature engineering and advanced class-imbalance handling techniques. The dataset, consisting of 494 HMH structures, is highly imbalanced across dimensionality classes (0D, 1D, 2D, 3D), posing significant challenges to predictive modeling. This dataset was later augmented to 1336 via the Synthetic Minority Oversampling Technique (SMOTE) to mitigate the effects of the class imbalance. We developed interaction-based descriptors and integrated them into a multi-stage workflow that combines feature selection, model stacking, and performance optimization to improve dimensionality prediction accuracy. Our approach significantly improves F1-scores for underrepresented classes, achieving robust cross-validation performance across all dimensionalities.

cs.LG