Korea Biochar Research Center, APRU Sustainable Waste Management & Division of Environmental Science and Ecological Engineering, Korea University, Seoul 02841, Republic of Korea; School of Environmental Science and Engineering, Guangzhou University, Guangzhou 510006, PR China.
The Sargent Centre for Process Systems Engineering, Department of Chemical Engineering, University College London, Torrington Place, London WC1E 7JE, UK.
Sci Total Environ. 2023 Dec 15;904:166678. doi: 10.1016/j.scitotenv.2023.166678. Epub 2023 Aug 31.
Arsenic (As) contamination in water is a significant environmental concern with profound implications for human health. Accurate prediction of the adsorption capacity of arsenite [As(III)] and arsenate [As(V)] on biochar is vital for the reclamation and recycling of polluted water resources. However, comprehending the intricate mechanisms that govern arsenic accumulation on biochar remains a formidable challenge. Data from the literature on As adsorption to biochar was compiled and fed into machine learning (ML) based modelling algorithms, including AdaBoost, LGBoost, and XGBoost, in order to build models to predict the adsorption efficiency of As(III) and As(V) to biochar, based on the compositional and structural properties. The XGBoost model showed superior accuracy and performance for prediction of As adsorption efficiency (for As(III): coefficient of determination (R) = 0.93 and root mean square error (RMSE) = 1.29; for As(V), R = 0.99, RMSE = 0.62). The initial concentrations of As(III) and As(V) as well as the dosage of the adsorbent were the most significant factors influencing adsorption, explaining 48 % and 66 % of the variability for As(III) and As(V), respectively. The structural properties and composition of the biochar explained 12 % and 40 %, respectively, of the variability of As(III) adsorption, and 13 % and 21 % of that of As(V). The XGBoost models were validated using experimental data. R values were 0.9 and 0.84, and RMSE values 6.5 and 8.90 for As(III) and As(V), respectively. The ML approach can be a valuable tool for improving the treatment of inorganic As in aqueous environments as it can help estimate the optimal adsorption conditions of As in biochar-amended water, and serve as an early warning for As-contaminated water.
砷（As）污染是一个重大的环境问题，对人类健康有着深远的影响。准确预测亚砷酸盐[As(III)]和砷酸盐[As(V)]在生物炭上的吸附容量对于受污染水资源的回收和再利用至关重要。然而，理解砷在生物炭上积累的复杂机制仍然是一个巨大的挑战。我们收集了关于砷在生物炭上吸附的文献数据，并将其输入到基于机器学习（ML）的建模算法中，包括 AdaBoost、LGBoost 和 XGBoost，以构建基于生物炭组成和结构特性预测 As(III)和 As(V)吸附效率的模型。XGBoost 模型在预测 As 吸附效率方面表现出更高的准确性和性能（对于 As(III)：决定系数 (R)为 0.93，均方根误差 (RMSE)为 1.29；对于 As(V)，R 为 0.99，RMSE 为 0.62）。As(III)和 As(V)的初始浓度以及吸附剂的剂量是影响吸附的最重要因素，分别解释了 As(III)和 As(V)变异性的 48%和 66%。生物炭的结构特性和组成分别解释了 As(III)吸附变异性的 12%和 40%，以及 As(V)吸附变异性的 13%和 21%。XGBoost 模型使用实验数据进行了验证。对于 As(III)和 As(V)，R 值分别为 0.9 和 0.84，RMSE 值分别为 6.5 和 8.90。ML 方法可以成为改善水中无机砷处理的有效工具，因为它可以帮助估计生物炭改良水中砷的最佳吸附条件，并作为受污染水的早期预警。
