收藏切换
Explainable machine learning for uncovering PFAS adsorption mechanisms in soils
收藏切换
Bing LIU1, *, Ke-ke ZOU2, Hui LIU1, Kai-ge SHI1, Yan-min CHEN1, Xin LU1, Shuai-jun HAN1, Li GU3
China Environmental Science | 2026, 46(3) : 1460 - 1474
Less
收藏切换
China Environmental Science | 2026, 46(3): 1460-1474
• Soil Pollution Control •
Explainable machine learning for uncovering PFAS adsorption mechanisms in soils
Full
Bing LIU1, *, Ke-ke ZOU2, Hui LIU1, Kai-ge SHI1, Yan-min CHEN1, Xin LU1, Shuai-jun HAN1, Li GU3
Affiliations
  • 1.School of Chemistry and Chemical Engineering, Zhengzhou Normal University, Zhengzhou 450044, China
  • 2.Pingdingshan Hydrology and Water Resources Reporting Subcenter of Henan Province, Pingdingshan 467000, China
  • 3.College of Environment and Ecology, Chongqing University, Chongqing 400044, China
Published: 2026-03-20
Outline
收藏切换

The soil-water partition coefficient (Kd) of per- and polyfluoroalkyl substances (PFAS) was predicted using machine learning (ML) models, and their partitioning behavior in soils was elucidated. A dataset comprising 1,227 samples of 47 PFAS was employed, with 16 PFAS physicochemical properties and soil parameters used as input features. It was demonstrated that the lgKd values ranged from -1.40 to 3.95. When the number of CF2 groups exceeded 5and the molecular weight (MW) was greater than 400g/mol, significant linear relationships were observed between lgKd and both the CF2 number and MW, with R2 values of 0.96 and 0.94, respectively. The influence of soil properties on PFAS adsorption was found to be dependent on the type of PFAS: non-zwitterionic PFAS were mainly affected by organic carbon content and silt content, whereas zwitterionic PFAS were primarily influenced by silt and sand content. After parameter optimization, the ML model exhibited good predictive performance, achieving an R2 of 0.85, an RMSE of 0.35, and an MAE of 0.26. MW, organic carbon content, water solubility (lgS), and net charge density were identified as the most important features, with contribution rates of 31.6%, 29.8%, 19.7%, and 10.0%, respectively. Three-dimensional interaction analysis indicated that when soils with high organic carbon content, high cation exchange capacity (CEC), and low pH were combined with PFAS of large MW, lgKd exceeded 0.40, reflecting strong adsorption. Conversely, under conditions where CEC was below 8.00 cmol/kg, organic matter content was less than 1.00%, and MW was below 380g/mol, lgKd was lower than 0.40, indicating weaker adsorption and higher potential for environmental migration. The findings provide a reliable tool for predicting the adsorption behavior of PFAS.

per-and polyfluoroalkyl substances  /  explainable machine learning  /  soil-water partition coefficient  /  prediction
刘冰, 邹可可, 刘辉, 时凯歌, 陈燕敏, 卢鑫, 韩帅军, 古励. 基于可解释性机器学习的PFAS土壤中分配机制. 中国环境科学, 2026 , 46 (3) : 1460 -1474 .
Bing LIU, Ke-ke ZOU, Hui LIU, Kai-ge SHI, Yan-min CHEN, Xin LU, Shuai-jun HAN, Li GU. Explainable machine learning for uncovering PFAS adsorption mechanisms in soils[J]. China Environmental Science, 2026 , 46 (3) : 1460 -1474 .
Year 2026 volume 46 Issue 3
93
18
Cite this Article
BibTeX
Article Info
  • Receive Date:2025-08-03
  • Online Date:2026-10-08
  • Published:2026-03-20
Article Data
Affiliations
History
  • Received:2025-08-03
Funding
Affiliations
    1.School of Chemistry and Chemical Engineering, Zhengzhou Normal University, Zhengzhou 450044, China
    2.Pingdingshan Hydrology and Water Resources Reporting Subcenter of Henan Province, Pingdingshan 467000, China
    3.College of Environment and Ecology, Chongqing University, Chongqing 400044, China
References
Share
https://castjournals.cast.org.cn/joweb/zghjkx/EN/
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT