Scholarlink

<p>This study investigates the potential applications of iron nanoparticles (FeNPs) synthesized by the green synthesis method in the dye removal of wastewater. 89 data points were obtained using various parameters (initial dye concentration, pH, temperature, contact time, and adsorbent dosage). The study examines how these parameters affect the removal efficiency. Each parameter varies within specific ranges, including pH (1–14), temperature (10–40&#xa0;°C), contact time (2–90&#xa0;min), initial pollutant concentration (1–750&#xa0;mg/L), and adsorbent dosage (0.01–2.0&#xa0;g/L). Optimal conditions are obtained as 11 for pH, initial dye concentration 100&#xa0;mg/L, temperature 25&#xa0;°C, contact time 90&#xa0;min, and adsorbent dosage 0.5&#xa0;g/L. In addition, artificial intelligence-based models are used to estimate the efficiency of dye extraction. Five different machine learning (ML) approaches have been evaluated within the scope of the study (Random Forest-RF, Decision Tree-DT, K-Nearest Neighbors (KNN), Linear Regression-LR, and Support Vector Regression-SVR). The RF model has shown high performance on the test data after Box-Cox transformation and hyperparameter optimization (R<sup>2</sup> = 0.945, root mean square error (RMSE) = 6.50, mean squared error (MSE) = 42.28, mean absolute error (MAE) = 4.29) and was determined to be a strong approach in terms of predictive power. The RF model used in this study can be used to estimate dye removal efficiency under untested conditions. This approach has the potential to provide significant savings in both time and financial resources by reducing experimental workload and facilitating the design of more efficient treatment processes.</p> Graphical abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimizing wastewater treatment: artificial intelligence-based prediction and green synthesis of iron nanoparticles for efficient dye removal

  • 期刊论文

  • 发表日期：

    2025年11月27日

查看全文
  • Ş. M. Yakut,
  • S. Atasever

摘要

This study investigates the potential applications of iron nanoparticles (FeNPs) synthesized by the green synthesis method in the dye removal of wastewater. 89 data points were obtained using various parameters (initial dye concentration, pH, temperature, contact time, and adsorbent dosage). The study examines how these parameters affect the removal efficiency. Each parameter varies within specific ranges, including pH (1–14), temperature (10–40 °C), contact time (2–90 min), initial pollutant concentration (1–750 mg/L), and adsorbent dosage (0.01–2.0 g/L). Optimal conditions are obtained as 11 for pH, initial dye concentration 100 mg/L, temperature 25 °C, contact time 90 min, and adsorbent dosage 0.5 g/L. In addition, artificial intelligence-based models are used to estimate the efficiency of dye extraction. Five different machine learning (ML) approaches have been evaluated within the scope of the study (Random Forest-RF, Decision Tree-DT, K-Nearest Neighbors (KNN), Linear Regression-LR, and Support Vector Regression-SVR). The RF model has shown high performance on the test data after Box-Cox transformation and hyperparameter optimization (R2 = 0.945, root mean square error (RMSE) = 6.50, mean squared error (MSE) = 42.28, mean absolute error (MAE) = 4.29) and was determined to be a strong approach in terms of predictive power. The RF model used in this study can be used to estimate dye removal efficiency under untested conditions. This approach has the potential to provide significant savings in both time and financial resources by reducing experimental workload and facilitating the design of more efficient treatment processes.

Graphical abstract