arXiv subjects

Shoeb Athar

Publications and source records attributed to Shoeb Athar.

5 recordsLinked to original sources

arXiv2026

An All-Atom Model of the Resorcinol-Formaldehyde Aerogel Skeleton from DFT and Molecular Dynamics: Validation and Iron Incorporation

Resorcinol-formaldehyde (RF) aerogels are lightweight, ultra-porous thermal superinsulators whose complex amorphous structure has not yet been described at the all-atom level. Here, we build an all-atom model of the RF aerogel skeleton by combining density functional theory (DFT) and classical molecular dynamics (MD), validate it against experimental infrared and density measurements, and use it to study the incorporation of iron dopants. The RF base unit was optimized with Gaussian, DMol$^{3}$ and SIESTA, and its calculated infrared spectrum agreed with experiment. Comparing classical force fields with DFT for branched systems of two to four base units (140-278 atoms) led to the choice of COMPASS II. A 15,012-atom system of 54 branched RF units, not cross-linked to each other, aggregated into a single cluster in the canonical (NVT) ensemble and reached an equilibrium density of 1.296 g.cm$^{-3}$ in the isobaric-isothermal (NPT) ensemble, in agreement with the experimental skeletal density (1.3 g.cm$^{-3}$). Then, 5 wt$\%$ iron (109 atoms) was added to the equilibrated structure: during 1 ns of NPT simulation, the fraction of iron atoms embedded in the skeleton rose from 63$\%$ to 71$\%$, and the Fe-C radial distribution function peaked at 3 Å. The model thus provides an experimentally supported starting point for screening fillers and dopants in RF gels, with prospective application to their thermoelectric functionalization, which we have previously demonstrated experimentally with basalt and PANOX fibres. Future extensions could address skeleton properties relevant to thermal insulation and the pyrolysis of RF gels into carbon and metal-doped carbon aerogels.

cond-mat.mtrl-sci↗

arXiv2026

A Generalizable Machine Learning Framework for Half-Heusler Thermoelectrics

Machine learning (ML) has emerged as a powerful approach for accelerating the discovery of thermoelectric (TE) materials, yet many models suffer from limited generalizability. Here we present a robust ML framework for predicting the thermoelectric figure of merit (zT) in half-Heusler (hH) compounds, emphasizing prediction of previously unseen compositions. A principal component analysis (PCA)-based train/test splitting strategy is introduced to generate representative partitions of chemical space. Bayesian hyperparameter optimization is combined with correlation-based k-best feature selection for Random Forest, XGBoost, and Neural Network models, together with the interpretable SISSO symbolic-regression method. Relative to the full set of 118 input features, the optimized feature subsets reduce the test RMSE by 16-27%. SHAP and SISSO analyses identify temperature, A-site dopant concentration, and A-site heat of vaporization as the most important descriptors of zT. The framework is validated through prediction of previously unexplored compositions, recovery of literature-reported optimal zT values for established hH systems, and first-principles calculations for the new compounds CeCoBi and CeIrBi, which exhibit favorable electronic transport properties. The framework is further applied to high-throughput screening of the doped hH chemical space, covering about 2.46 million elemental substitution combinations across 631 thermodynamically screened parent matrices under stability constraints. This screening yields a concise set of promising candidates for future study. Overall, the work establishes a generalizable workflow for reliable data-driven thermoelectric materials discovery.

cond-mat.mtrl-sci↗

arXiv2026

Beyond Predicted ZT: Machine Learning Strategies for the Experimental Discovery of Thermoelectric Materials

The discovery of high-performance thermoelectric (TE) materials for advancing green energy harvesting from waste heat is an urgent need in the context of looming energy crisis and climate change. The rapid advancement of machine learning (ML) has accelerated the design of thermoelectric (TE) materials, yet a persistent "gap" remains between high-accuracy computational predictions and their successful experimental validation. While ML models frequently report impressive test scores (R^2 values of 0.90-0.98) for complex TE properties (zT, power factor, and electrical/thermal conductivity), only a handful of these predictions have culminated in the experimental discovery of new high-zT materials. In this review, we identify and discuss that the primary obstacles are poor model generalizability-stemming from the "small-data" problem, sampling biases in cross-validation, and inadequate structural representation-alongside the critical challenge of thermodynamic phase stability. Moreover, we argue that standard randomized validation often overestimates model performance by ignoring "hidden hierarchies" and clustering within chemical families. Finally, to bridge this gap between ML-predictions and experimental realization, we advocate for advanced validation strategies like PCA-based sampling and a synergetic active learning loop that integrates ML "fast filters" for stability (e.g., GNoME) with high-throughput combinatorial thin-film synthesis to rapidly map stable, high-zT compositional spaces.

cond-mat.mtrl-sci↗

arXiv2025

Tackling dataset curation challenges towards reliable machine learning: a case study on thermoelectric materials

Machine Learning (ML) driven discovery of novel and efficient thermoelectric (TE) materials warrants experimental TE datasets of high volume, diversity, and quality. While the largest publicly available dataset, Starrydata2, has a high data volume, it contains inaccurate data due to the inherent limitations of Large Language Model (LLM)-assisted data curation, ambiguous nomenclature and complex formulas of materials in the literature. Another unaddressed issue is the inclusion of multi-source experimental data, with high standard deviations and without synthesis information. Using half-Heusler (hH) materials as an example, this work is aimed at first highlighting these errors and inconsistencies which cannot be filtered with conventional dataset curation workflows. We then propose a statistical round-robin error-based data filtering method to address these issues, a method that can be applied to filter any other material property. Lastly, a hybrid dataset creation workflow, involving data from Starrydata2 and manual extraction, is proposed and the resulting dataset is analyzed and compared against Starrydata2.

cond-mat.mtrl-sci↗

arXiv2025

Carbogels for sustainable and scalable thermoelectric applications

Thermoelectric generators (TEGs) based on commercially used thermal super-insulating materials can facilitate sustainable and large-scale ambient waste heat recovery while bequeathing an added economic and environmental value to thermal insulations in industry. This requires the optimization of the thermoelectric (TE) properties through electrical functionalization of such materials. Moreover, the associated engineering challenges of assembling TEG modules must be overcome. Herein, we propose using super-insulating Resorcinol-formaldehyde (RF) carbogels for scalable and sustainable TE applications through their electrical functionalization. Using a combination of a pyrolysis process and carbon fibers insertion, we achieved an increment by 12 orders of magnitude in electrical conductivity as well as ZT whilst retaining their intrinsic ultralow thermal conductivity (<50 mW/mK). A TE module in the form of a thermoelectric vacuum insulation panel (TVIP), was then fabricated using only a p-type material, to demonstrate a proof-of-concept self-powered WiFi-based vacuum-failure detection application in confined spaces in automobiles or aeronautics. Finally, by extrapolating the optimized output power and with a CAD-assisted assembly of a large TEG module (1000 cm2), the potential of scalable low-grade waste heat recovery is discussed.

physics.app-ph↗