Search for: All records

Abstract Ensuring the long-term chemical durability of glasses is critical for nuclear waste immobilization operations. Durable glasses usually undergo qualification for disposal based on their response to standardized tests such as the product consistency test or the vapor hydration test (VHT). The VHT uses elevated temperature and water vapor to accelerate glass alteration and the formation of secondary phases. Understanding the relationship between glass composition and VHT response is of fundamental and practical interest. However, this relationship is complex, non-linear, and sometimes fairly variable, posing challenges in identifying the distinct effect of individual oxides on VHT response. Here, we leverage a dataset comprising 654 Hanford low-activity waste (LAW) glasses across a wide compositional envelope and employ various machine learning techniques to explore this relationship. We find that Gaussian process regression (GPR), a nonparametric regression method, yields the highest predictive accuracy. By utilizing the trained model, we discern the influence of each oxide on the glasses’ VHT response. Moreover, we discuss the trade-off between underfitting and overfitting for extrapolating the material performance in the context of sparse and heterogeneous datasets. 

Abstract Portlandite (Ca(OH)₂; also known as calcium hydroxide or hydrated lime), an archetypal alkaline solid, interacts with carbon dioxide (CO₂) via a classic acid–base “carbonation” reaction to produce a salt (calcium carbonate: CaCO₃) that functions as a low‐carbon cementation agent, and water. Herein, we revisit the effects of reaction temperature, relative humidity (RH), and CO₂concentration on the carbonation of portlandite in the form of finely divided particulates and compacted monoliths. Special focus is paid to uncover the influences of the moisture state (i.e., the presence of adsorbed and/or liquid water), moisture content and the surface area‐to‐volume ratio (s_a/v, mm⁻¹) of reactants on the extent of carbonation. In general, increasing RH more significantly impacts the rate and thermodynamics of carbonation reactions, leading to high(er) conversion regardless of prior exposure history. This mitigated the effects (if any) of allegedly denser, less porous carbonate surface layers formed at lower RH. In monolithic compacts, microstructural (i.e., mass‐transfer) constraints particularly hindered the progress of carbonation due to pore blocking by liquid water in compacts with limited surface area to volume ratios. These mechanistic insights into portlandite's carbonation inform processing routes for the production of cementation agents that seek to utilize CO₂borne in dilute (≤30 mol%) post‐combustion flue gas streams. 

It is significant to investigate the calcium carbonate (CaCO3) precipitation mechanism during the carbon capture process; nevertheless, CaCO3 precipitation is not clearly understood yet. Understanding the carbonation mechanism at the atomic level can contribute to the mineralization capture and utilization of carbon dioxide, as well as the development of new cementitious materials with high-performance. There are many factors, such as temperature and CO2 concentration, that can influence the carbonation reaction. In order to achieve better carbonation efficiency, the reaction conditions of carbonation should be fully verified. Therefore, based on molecular dynamics simulations, this paper investigates the atomic-scale mechanism of carbonation. We investigate the effect of carbonation factors, including temperature and concentration, on the kinetics of carbonation (polymerization rate and activation energy), the early nucleation of calcium carbonate, etc. Then, we analyze the local stresses of atoms to reveal the driving force of early stage carbonate nucleation and the reasons for the evolution of polymerization rate and activation energy. Results show that the higher the calcium concentration or temperature, the higher the polymerization rate of calcium carbonate. In addition, the activation energies of the carbonation reaction increase with the decrease in calcium concentrations. 

Our multi-task neural network approach simultaneously predicts the concentration of all types of rare earth elements (REEs) in coal ashes, with an improved accuracy and robustness as compared to conventional single-task neural networks. 

Architected materials design across orders of magnitude length scale intrigues exceptional mechanical responses nonexistent in their natural bulk state. However, the so‐termed mechanical metamaterials, when scaling bottom down to the atomistic or microparticle level, remain largely unexplored and conventionally fall out of their coarse‐resolution, ordered‐pattern design space. Here, combining high‐throughput molecular dynamics (MD) simulations and machine learning (ML) strategies, some intriguing atomistic families of disordered mechanical metamaterials are discovered, as fabricated by melt quenching and exemplified herein by lightweight‐yet‐stiff cellular materials featuring a theoretical limit of linear stiffness–density scaling, whose structural disorder—rather than order—is key to reduce the scaling exponent and is simply controlled by the bonding interactions and their directionality that enable flexible tunability experimentally. Importantly, a systematic navigation in the forcefield landscape reveals that, in‐between directional and non‐directional bonding such as covalent and ionic bonds, modest bond directionality is most likely to promotes disordered packing of polyhedral, stretching‐dominated structures responsible for the formation of metamaterials. This work pioneers a bottom‐down atomistic scheme to design mechanical metamaterials formatted disorderly, unlocking a largely untapped field in leveraging structural disorder in devising metamaterials atomistically and, potentially, generic to conventional upscaled designs. 

Numerical simulations have revolutionized material design. However, although simulations excel at mapping an input material to its output property, their direct application to inverse design has traditionally been limited by their high computing cost and lack of differentiability. Here, taking the example of the inverse design of a porous matrix featuring targeted sorption isotherm, we introduce a computational inverse design framework that addresses these challenges, by programming differentiable simulation on TensorFlow platform that leverages automated end-to-end differentiation. Thanks to its differentiability, the simulation is used to directly train a deep generative model, which outputs an optimal porous matrix based on an arbitrary input sorption isotherm curve. Importantly, this inverse design pipeline leverages the power of tensor processing units (TPU)—an emerging family of dedicated chips, which, although they are specialized in deep learning, are flexible enough for intensive scientific simulations. This approach holds promise to accelerate inverse materials design. 

Calcium–alumino–silicate–hydrate (CaO–Al2O3–SiO2–H2O, or C–A–S–H) gel, which is the binding phase of cement-based materials, greatly influences concrete mechanical properties and durability. However, the atomic-scale kinetics of the aluminosilicate network condensation remains puzzling. Here, based on reactive molecular dynamics simulations of C–A–S–H systems formation with varying Al/Ca molar ratios, we study the kinetic mechanism of the hydrated aluminosilicate gels upon precipitation. We show that the condensation activation energy decreases with the Al/Ca molar ratio, which suggests that the concentration of the Al polytopes has a great effect on controlling the kinetics of the gelation reaction. Significantly, we demonstrate that 5-fold Al atoms are mainly forming at high Al/Ca molar ratios since there are insufficient hydrogen cations or extra calcium cations to compensate the negatively charged Al polytopes at high Al/Ca molar ratios during accelerated aging. 

A graph-based machine learning model is built to predict atom dynamics from their static structure, which, in turn, unveils the predictive power of static structure in dynamical evolution of disordered phases.