Machine learning for layer-by-layer nanofiltration membrane performance prediction and polymer candidate exploration

Chen Wang; Li Wang; Hanwei Yu; Allan Seo; Zhining Wang; Saeid Rajabzadeh; Bing-Jie Ni; Ho Kyong Shon

doi:10.1016/j.chemosphere.2023.140999

Machine learning for layer-by-layer nanofiltration membrane performance prediction and polymer candidate exploration

Chemosphere. 2024 Feb:350:140999. doi: 10.1016/j.chemosphere.2023.140999. Epub 2023 Dec 25.

Authors

Chen Wang¹, Li Wang², Hanwei Yu¹, Allan Seo¹, Zhining Wang³, Saeid Rajabzadeh¹, Bing-Jie Ni⁴, Ho Kyong Shon⁵

Affiliations

¹ School of Civil and Environmental Engineering, University of Technology Sydney, Sydney, New South Wales, 2007, Australia.
² CSIRO Space and Astronomy, PO Box 1130, Bentley, WA, 6102, Australia.
³ Shandong Provincial Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science and Engineering, Shandong University, Qingdao, 266237, China.
⁴ School of Civil and Environmental Engineering, University of New South Wales, Sydney, New South Wales, 2052, Australia.
⁵ School of Civil and Environmental Engineering, University of Technology Sydney, Sydney, New South Wales, 2007, Australia. Electronic address: Hokyong.Shon-1@uts.edu.au.

PMID: 38151066
DOI: 10.1016/j.chemosphere.2023.140999

Abstract

In this study, machine learning-based models were established for layer-by-layer (LBL) nanofiltration (NF) membrane performance prediction and polymer candidate exploration. Four different models, i.e., linear, random forest (RF), boosted tree (BT), and eXtreme Gradient Boosting (XGBoost), were formed, and membrane performance prediction was determined in terms of membrane permeability and selectivity. The XGBoost exhibited optimal prediction accuracy for membrane permeability (coefficient of determination (R²): 0.99) and membrane selectivity (R²: 0.80). The Shapley Additive exPlanation (SHAP) method was utilized to evaluate the effects of different LBL NF membrane fabrication conditions on membrane performances. The SHAP method was also used to identify the relationships between polymer structure and membrane performance. Polymers were represented by Morgan fingerprint, which is an effective description approach for developing modeling. Based on the SHAP value results, two reference Morgan fingerprints were constructed containing atomic groups with positive contributions to membrane permeability and selectivity. According to the reference Morgan fingerprint, 204 potential polymers were explored from the largest polymer database (PoLyInfo). By calculating the similarities between each potential polymer and both reference Morgan fingerprints, 23 polymer candidates were selected and could be further used for LBL NF membrane fabrication with the potential for providing good membrane performance. Overall, this work provided new ways both for LBL NF membrane performance prediction and high-performance polymer candidate exploration. The source code for the models and algorithms used in this study is publicly available to facilitate replication and further research. https://github.com/wangliwfsd/LLNMPP/.

Keywords: Layer-by-layer membrane; Machine learning; Nanofiltration; Permeability; Polymer; Selectivity.

MeSH terms

Algorithms*
Databases, Factual
Machine Learning*
Membranes
Polymers

Substances

Polymers