Discovery of an End-to-End Pattern for Contaminant-Oriented Advanced Oxidation Processes Catalyzed by Biochar with Explainable Machine Learning

Environmental Science & Technology, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 20, 2024

The utilization of biochar-catalyzed peroxymonosulfate in advanced oxidation processes (BC-PMS AOPs) is widely acknowledged as an effective and economical method for mitigating emerging contaminants (ECs). Especially, state-of-the-art machine learning (ML) technology has been employed to accurately predict the reaction rate constants EC degradation BC-PMS AOPs, primarily focusing on three aspects: performance prediction, operating condition optimization, mechanism interpretation. However, its real application specific optimization targeting different ECs seldom considered, hindering realization contaminant-oriented AOPs. Herein, we propose a hierarchical ML pipeline achieve end-to-end (E2E) pattern addressing this issue. First, overall XGB model, trained with comprehensive data set, can perform well predicting additionally providing basis further analysis various ECs. Then, submodels clusters offer strategies selection optimum option AOPs HOMO-LUMO gaps, thus forming E2E This study not only increases our understanding but also successfully bridges gap between model development environmental application.

Language: Английский

---

Interpretable Machine Learning Models Delivering a New Perspective for the Reaction Mechanism between Organic Pollutants and Oxidative Radicals

Environmental Science & Technology, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 7, 2025

Machine learning (ML) is expected to bring new insights into the impact of organic structures on reaction mechanisms in reactive oxygen species oxidation. However, understanding underlying chemical still faces challenges due limited interpretability ML models. In this study, interpretable models were established predict second-order rate constants between hydroxyl radicals (•OH) and organics (k•OH). It was found that energy highest occupied molecular orbital (EHOMO), number aromatic rings (NAR), carbon atoms (NC) have important impacts k•OH. The positive correlation k•OH EHOMO can be explained by regularity electrophilic reaction, while relationship NAR NC seems related with sites. Furthermore, a rapid judgment method for mechanism developed based an unsupervised approach which automatically divided three clusters. Additionally, methodology applied sulfate radicals. This study offers rational model predicting provides more from perspective big data.

Language: Английский

---

Gradient boosting decision tree algorithms for accelerating nanofiltration membrane design and discovery

Desalination, Journal Year: 2024, Volume and Issue: 592, P. 118072 - 118072

Published: Sept. 2, 2024

Language: Английский

---

Predicting the Performance of Lithium Adsorption and Recovery from Unconventional Water Sources with Machine Learning

Water Research, Journal Year: 2024, Volume and Issue: 266, P. 122374 - 122374

Published: Sept. 7, 2024

Language: Английский

---

Efficient electrochemical oxidation of ofloxacin by IrO2 -RuO2-TiO2 /Ti anode: Parameters optimization, kinetics and degradation pathways

Environmental Pollution, Journal Year: 2025, Volume and Issue: unknown, P. 126216 - 126216

Published: April 1, 2025

Language: Английский

---

Predictive modeling of Enterococcus sp. removal with limited data from different advanced oxidation processes: A machine learning approach

Journal of environmental chemical engineering, Journal Year: 2024, Volume and Issue: 12(3), P. 112530 - 112530

Published: March 19, 2024

The removal of contaminants through Advanced Oxidation Processes (AOPs) is a complex task that demands the simultaneous consideration multiple operating parameters, such as type and concentration oxidant catalyst, intensity radiation, composition aqueous matrix, etc. Designing efficient AOPs often requires expensive time-consuming laboratory experiments. To improve this process, study proposes Machine Learning approach based on Random Forest (RF) model, to predict Enterococcus sp. in wastewater treated with various AOPs, even when dealing limited data. assess our under diverse conditions, data partitioning methodology used categorize different into three distinct cases increasing complexity, from Case I III. evaluation RF model's performance, combined methodology, demonstrated its usefulness predicting missing or additional disinfection values at any instant during AOPs. Specifically, I, model excels generalizing predictions across AOP treatments, followed by II III, which achieve Root Mean Squared Error (RMSE) below comparable average RMSE (0.72) 8 out 15 2 4 respectively. Moreover, effects imbalanced performance are discussed. This highlights potential facilitate design new experiments same treatment without need for trials, challenging conditions.

Language: Английский

---

Deciphering N-Doped Biochar Design for Non-Radical Pathways through Hierarchical Machine Learning

ACS ES&T Engineering, Journal Year: 2024, Volume and Issue: 4(7), P. 1738 - 1747

Published: April 3, 2024

Biochar has been widely employed for the promotion of advanced oxidation processes (AOPs) and when combined with nitrogen doping charge distribution mediation, N-doped biochar (NBC) can serve as a highly effective catalyst degradation persistent organic pollutants. However, due to variety preparation methods, intrinsic active sites AOP catalysis have not clearly identified. Furthermore, complex relationships between method, material properties, catalytic pathways remain unclear, impeding widespread practical application NBC. Herein, machine learning (ML) was implemented predict pathway identify vital properties N-doping required acceleration AOPs. During process model training, an innovative method data set splitting applied, comparing results generated from multiple models enhance interpretability. We elucidated correlation primary features nonradical pathway, focusing on contribution N species regulatory role pyrolysis temperature. Detailed insights were further provided ratio design NBC mediation. Overall, this study offers novel into NBC-mediated AOPs pollution control, underscoring significant potential ML accelerating applications.

Language: Английский

---

Task Decomposition Strategy Based on Machine Learning for Boosting Performance and Identifying Mechanisms in Heterogeneous Activation of Peracetic Acid Process

Water Research, Journal Year: 2024, Volume and Issue: 267, P. 122521 - 122521

Published: Sept. 26, 2024

Language: Английский

---

Improving prediction of N2O emissions during composting using model-agnostic meta-learning

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 922, P. 171357 - 171357

Published: Feb. 29, 2024

Language: Английский

---

Autoencoded chemical feature interaction machine learning method boosting performance of piezoelectric catalytic process

Nano Energy, Journal Year: 2024, Volume and Issue: 126, P. 109670 - 109670

Published: April 23, 2024

Language: Английский

---