Molecular simulations elucidate the transport mechanisms of organic solvents in dense polymer membranes

Journal of Membrane Science, Год журнала: 2024, Номер 708, С. 123055 - 123055

Опубликована: Июль 3, 2024

Язык: Английский

---

Gas permeability, diffusivity, and solubility in polymers: Simulation-experiment data fusion and multi-task machine learning

npj Computational Materials, Год журнала: 2024, Номер 10(1)

Опубликована: Авг. 15, 2024

Machine learning (ML) models for predicting gas permeability through polymers have traditionally relied on experimental data. While these exhibit robustness within familiar chemical domains, reliability wanes when applied to new spaces. To address this challenge, we present a multi-tiered multi-task framework empowered with advanced machine-crafted polymer fingerprinting algorithms and data fusion techniques. This combines scarce "high-fidelity" abundant diverse "low-fidelity" simulation or synthetic data, resulting in predictive that display high level of generalizability across novel Additionally, scheme capitalizes known physics interrelated properties, such as diffusivity solubility, both which are closely tied permeability. By amalgamating throughput generated available permeability, diffusivity, solubility various gases, construct deep models. These can simultaneously predict all three properties gases under consideration, markedly enhanced accuracy, particularly compared traditional reliant solely singular property. strategy underscores the potential coupling high-throughput classical simulations methodologies yield state-of-the-art property predictors, especially targeted is scarce.

Язык: Английский

---

Precise prediction of CO2 separation performance of metal–organic framework mixed matrix membranes based on feature selection and machine learning

Separation and Purification Technology, Год журнала: 2024, Номер 349, С. 127894 - 127894

Опубликована: Май 9, 2024

Язык: Английский

---

Machine Learning for Polymer Design to Enhance Pervaporation-Based Organic Recovery

Environmental Science & Technology, Год журнала: 2024, Номер 58(23), С. 10128 - 10139

Опубликована: Май 14, 2024

Pervaporation (PV) is an effective membrane separation process for organic dehydration, recovery, and upgrading. However, it crucial to improve materials beyond the current permeability-selectivity trade-off. In this research, we introduce machine learning (ML) models identify high-potential polymers, greatly improving efficiency reducing cost compared conventional trial-and-error approach. We utilized largest PV data set date incorporated polymer fingerprints features, including structure, operating conditions, solute properties. Dimensionality reduction, missing treatment, seed randomness, leakage management were employed ensure model robustness. The optimized LightGBM achieved RMSE of 0.447 0.360 factor total flux, respectively (logarithmic scale). Screening approximately 1 million hypothetical polymers with ML resulted in identifying a predicted permeation index >30 synthetic accessibility score <3.7 acetic acid extraction. This study demonstrates promise accelerate tailored designs.

Язык: Английский

---

Predictive transport modelling in polymeric gas separation membranes: From additive contributions to machine learning

Separation and Purification Technology, Год журнала: 2024, Номер 340, С. 126743 - 126743

Опубликована: Фев. 12, 2024

Язык: Английский

---