<p>This study investigates the adsorption of Methyl Red (MR) dye from seawater using a novel biosorbent synthesized from zinc oxide-modified <i>Ligustrum japonicum</i> fruit (FLj/ZnO). The adsorbent was synthesized through a green method involving mixing a 0.2&#xa0;M ZnO solution with FLj fruit extract in a 2:1 (v/v) extract-to-solution ratio under controlled conditions (80&#xa0;°C, pH 10). Adsorption tests were conducted under optimized conditions: initial MR concentration = 20&#xa0;mg/L, FLj/ZnO dose = 100&#xa0;mg/L, solution volume = 10&#xa0;mL, contact time = 60&#xa0;min, and pH = 4. Characterization revealed homogeneously distributed ZnO nanoparticles with a point of zero charge (pHpzc) of 6.8. The process followed pseudo-first-order kinetics (R<sup>2</sup> &gt; 0.99) and fit the Langmuir isotherm model (R<sup>2</sup> = 0.991), with a maximum adsorption capacity of 204.77&#xa0;mg/g. Thermodynamic parameters confirmed a spontaneous (ΔG° &lt; 0) and exothermic (ΔH° = –150.52&#xa0;kJ/mol) process, with negative entropy change (ΔS° = –473.17&#xa0;J/mol·K) indicating increased order at the interface. A Gaussian Process Regression model with Bootstrap Aggregation (GPR_Bag) provided highly accurate prediction of adsorption capacity, with R = 0.9998 and RMSE &lt; 1.1. A user-friendly MATLAB interface was also developed for real-time adsorption performance prediction. These findings highlight the high adsorption potential and predictive reliability of the FLj/ZnO biosorbent, making it a promising, eco-friendly solution for industrial wastewater treatment.</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

Gaussian process regression with bootstrap aggregation for real-time prediction and optimization of methyl red adsorption

  • H. Khelili,
  • H. Tahraoui,
  • A. Amrane,
  • C. Sedrati,
  • K. G. Konan,
  • H. Atout,
  • A. F. Belhadj,
  • L. Guechi,
  • M. Guellal,
  • A. A. Assadi,
  • W. Elfellah,
  • L. Khezami,
  • M. Hjiri,
  • J. Zhang

摘要

This study investigates the adsorption of Methyl Red (MR) dye from seawater using a novel biosorbent synthesized from zinc oxide-modified Ligustrum japonicum fruit (FLj/ZnO). The adsorbent was synthesized through a green method involving mixing a 0.2 M ZnO solution with FLj fruit extract in a 2:1 (v/v) extract-to-solution ratio under controlled conditions (80 °C, pH 10). Adsorption tests were conducted under optimized conditions: initial MR concentration = 20 mg/L, FLj/ZnO dose = 100 mg/L, solution volume = 10 mL, contact time = 60 min, and pH = 4. Characterization revealed homogeneously distributed ZnO nanoparticles with a point of zero charge (pHpzc) of 6.8. The process followed pseudo-first-order kinetics (R2 > 0.99) and fit the Langmuir isotherm model (R2 = 0.991), with a maximum adsorption capacity of 204.77 mg/g. Thermodynamic parameters confirmed a spontaneous (ΔG° < 0) and exothermic (ΔH° = –150.52 kJ/mol) process, with negative entropy change (ΔS° = –473.17 J/mol·K) indicating increased order at the interface. A Gaussian Process Regression model with Bootstrap Aggregation (GPR_Bag) provided highly accurate prediction of adsorption capacity, with R = 0.9998 and RMSE < 1.1. A user-friendly MATLAB interface was also developed for real-time adsorption performance prediction. These findings highlight the high adsorption potential and predictive reliability of the FLj/ZnO biosorbent, making it a promising, eco-friendly solution for industrial wastewater treatment.