Informative Training Data for Efficient Property Prediction in Metal-Organic Frameworks by Active Learning DOI Creative Commons
Ashna Jose, Emilie DEVIJVER, N. Jakse

et al.

Published: Dec. 1, 2023

In recent data-driven approaches to materials discov- ery, scenarios where target quantities are expensive compute or measure often overlooked. such cases, it becomes imperative construct a training set that includes the most diverse, representative, and informative samples. Here, novel regression tree-based active learning algorithm is employed for purpose. It applied predict band gap adsorption properties of metal-organic frameworks (MOFs), class results from virtually infinite combinations their building units. Simpler low dimensional descrip- tors, as Stoichiometric-120 geometric properties, found here better represent MOFs in data regime, used feature space this model. The partition given by tree constructed on labeled part dataset select new samples be added set, thereby limiting its size while maximizing prediction quality. Through tests QMOF, hMOF, dMOF sets, we show our method effective constructing small sets learn models well thus reducing label- ing cost. Specifically, approach highly beneficial when labels unevenly distributed descriptor label distribution imbalanced, which case real world data. This offers unique tool efficiently analyze complex structure-property relationships accelerate discovery.

Language: Английский

Machine learning-based prediction and experimental validation of heavy metal adsorption capacity of bentonite DOI
Lisheng Guo, Xin Xu,

Cencen Niu

et al.

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 926, P. 171986 - 171986

Published: March 28, 2024

Language: Английский

Citations

27
Machine learning for CO2 capture and conversion: A review DOI Creative Commons
Sung Eun Jerng, Yang Jeong Park, Ju Li

et al.

Energy and AI, Journal Year: 2024, Volume and Issue: 16, P. 100361 - 100361

Published: March 30, 2024

Coupled electrochemical systems for the direct capture and conversion of CO2 have garnered significant attention owing to their potential enhance energy- cost-efficiency by circumventing amine regeneration step. However, optimizing coupled system is more challenging than handling separated because its complexity, caused incorporation solvent heterogeneous catalysts. Nevertheless, deployment machine learning can be immensely beneficial, reducing both time cost ability simulate describe complex with numerous parameters involved. In this review, we summarized techniques employed in development solvents such as ionic liquids, well To optimize a system, these two separately developed will need combined via future.

Language: Английский

Citations

16
Informative Training Data for Efficient Property Prediction in Metal–Organic Frameworks by Active Learning DOI
Ashna Jose, Emilie DEVIJVER, N. Jakse

et al.

Journal of the American Chemical Society, Journal Year: 2024, Volume and Issue: 146(9), P. 6134 - 6144

Published: Feb. 26, 2024

In recent data-driven approaches to material discovery, scenarios where target quantities are expensive compute and measure often overlooked. such cases, it becomes imperative construct a training set that includes the most diverse, representative, informative samples. Here, novel regression tree-based active learning algorithm is employed for purpose. It applied predict band gap adsorption properties of metal-organic frameworks (MOFs), class materials results from virtually infinite combinations their building units. Simpler low dimensional descriptors, as those based on stoichiometric geometric properties, used feature space this model owing ability better represent MOFs in data regime. The partitions given by tree constructed labeled part select new samples be added set, thereby limiting its size while maximizing prediction quality. Tests QMOF, hMOF, dMOF sets reveal our method constructs small learn models more efficiently than existing approaches, with lower variance. Specifically, approach highly beneficial when labels unevenly distributed descriptor label distribution imbalanced, which case real world data. regions defined help revealing patterns data, offering unique tool analyze complex structure-property relationships accelerate discovery.

Language: Английский

Citations

12
Challenges and solutions to the scale-up of porous materials DOI Creative Commons
Marziyeh Nazari, Farnaz Zadehahmadi, Muhammad Sadiq

et al.

Communications Materials, Journal Year: 2024, Volume and Issue: 5(1)

Published: Aug. 29, 2024

With increasing pace, crystalline open frameworks are moving to larger scale, mature applications that stretch as broadly catalysis, separation, water purification, adsorption, sensing, biomineralization and energy storage. A particular challenge in this development can be the unexpected variation material properties from batch batch, even when a cursory analysis would indicate no process changes occurred. Our team has lived journey many projects where pilot scale production of metal-organic for use devices been key milestone suffered difficulties performance departures. In Perspective, we aim share some learning outcomes hope it will further speed field. major materials scale-up is between batches. Here, pilot-scale discussed suggestions provided help improve large-scale synthesis development.

Language: Английский

Citations

10
Control of Permanent Porosity in Type 3 Porous Liquids via Solvent Clustering DOI
Dennis Robinson Brown, Matthew J. Hurlock, Tina M. Nenoff

et al.

ACS Applied Materials & Interfaces, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 9, 2025

Porous liquids (PLs) are an exciting new class of materials for carbon capture due to their high gas adsorption capacity and ease industrial implementation. They composed sorbent particles suspended in a nonadsorbed solvent, forming liquid with permanent porosity. While PLs have vast number potential compositions based on the solvents available, most research has been focused selection rather than solvent. Therefore, PL design criteria supramolecular structures solvent explored create fundamental understanding how enables formation rapid discovery compositions. Atomistic molecular dynamics simulation eight range sizes, shapes, intramolecular bonding was performed, identifying that shape size clusters formed driving predictor individual molecule. The results demonstrate significant departure from common approaches steric exclusion molecules via pore aperture. A modeling experimental validation study further supports these findings. Through this computational material study, previously unexplored mechanism formation, solvent–solvent clustering, is identified as critical factor accelerated phase materials.

Language: Английский

Citations

1
Machine Learning for Gas Adsorption in Metal–Organic Frameworks: A Review on Predictive Descriptors DOI Creative Commons
I-Ting Sung,

Y. S. Cheng,

Chieh‐Ming Hsieh

et al.

Industrial & Engineering Chemistry Research, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 15, 2025

This review addresses a critical gap in the literature by focusing on features (or descriptors) used machine learning (ML) studies to predict gaseous adsorption properties metal–organic frameworks (MOFs). Although ML approaches for predicting MOFs have been extensively reported recent years, employed models not thoroughly reviewed. A comprehensive of these is crucial since they form foundation building effective predictive models. These are also key facilitating inverse design MOFs, as can be efficiently performance material candidates and explore structure–property relationship, guiding creation optimal MOF structures. Furthermore, naturally approaches, such encoder–decoder architectures. starts with brief overview importance applications various fields, followed discussion historical milestones computational research, highlighting role ML. then discusses traditional introduces newly proposed distinctive features, referred "beyond features", that date. generalized different gases outlined. Finally, we offer future outlooks ML-assisted searches applications. Overall, this aims help researchers grasp current developments insights into directions area.

Language: Английский

Citations

1
Machine Learning Descriptors for CO2 Capture Materials DOI Creative Commons

Ibrahim Orhan,

Yuankai Zhao, Ravichandar Babarao

et al.

Molecules, Journal Year: 2025, Volume and Issue: 30(3), P. 650 - 650

Published: Feb. 1, 2025

The influence of machine learning (ML) on scientific domains continues to grow, and the number publications at intersection ML, CO2 capture, material science is growing rapidly. Approaches for building ML models vary in both objectives methods through which materials are represented (i.e., featurised). Featurisation based descriptors, being a crucial step models, focus this review. Metal organic frameworks, ionic liquids, other discussed paper with descriptors used representation CO2-capturing materials. It shown that operating conditions must be included multiple temperatures and/or pressures used. Material can differentiate capture candidates falling under broad categories charge orbital, thermodynamic, structural, chemical composition-based descriptors. Depending application, dataset, model used, these carry varying degrees importance predictions made. Design strategies then derived selection important features. Overall, review predicts will play an even greater role future innovations capture.

Language: Английский

Citations

1
Carbon dioxide capturing activities of porous metal-organic frameworks (MOFs) DOI
Brij Mohan,

Virender Virender,

Ritika Kadiyan

et al.

Microporous and Mesoporous Materials, Journal Year: 2023, Volume and Issue: 366, P. 112932 - 112932

Published: Dec. 5, 2023

Language: Английский

Citations

20
Accelerated Discovery of Metal–Organic Frameworks for CO2 Capture by Artificial Intelligence DOI Creative Commons
Hasan Can Gülbalkan, Gokhan Onder Aksu, Goktug Ercakir

et al.

Industrial & Engineering Chemistry Research, Journal Year: 2023, Volume and Issue: 63(1), P. 37 - 48

Published: Dec. 25, 2023

The existence of a very large number porous materials is great opportunity to develop innovative technologies for carbon dioxide (CO2) capture address the climate change problem. On other hand, identifying most promising adsorbent and membrane candidates using iterative experimental testing brute-force computer simulations challenging due enormous variety materials. Artificial intelligence (AI) has recently been integrated into molecular modeling materials, specifically metal–organic frameworks (MOFs), accelerate design discovery high-performing adsorbents membranes CO2 adsorption separation. In this perspective, we highlight pioneering works in which AI, simulations, experiments have combined produce exceptional MOFs MOF-based composites that outperform traditional capture. We outline future directions by discussing current opportunities challenges field harnessing experiments, theory, AI accelerated

Language: Английский

Citations

14
Accelerated Discovery of CO2 Solid Sorbents Using Active Machine Learning: Review and Perspectives DOI

Deyang Xu,

Jing Yang,

Zhaoxiang Xu

et al.

Energy & Fuels, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 31, 2024

With the escalating severity of global climate change, significance carbon capture technology has become increasingly evident with respect to aim reaching peak and neutrality. Due exceptional selectivity, high adsorption capacity, long-term stability, solid sorbents are regarded as crucial materials for effective CO2 capture. Machine learning, an emerging tool in artificial intelligence, been adopted high-efficient screen catalysts recent years. By analyzing available data on material properties, machine learning can greatly enhance effectiveness precision identifying high-efficiency sorbents. This work provides overview latest advancements application capture, which specifically focuses by Several techniques their applications different types fully summarized concise comments, followed conclusion some challenges perspectives. review serve a guide development facilitate extensive utilization environmental protection.

Language: Английский

Citations

4