Scholarlink

<p>This study presents a novel integrated approach for deep desulfurization of both model and real diesel fuels through the combined development of a hybrid oscillatory-flow basket reactor, a newly synthesized magnetic heteropolymetalate carbon-based catalyst, and artificial intelligence–based predictive modeling. The proposed reactor design integrates key features of conventional basket reactors and central oscillatory-baffled reactors, forming centrally baffled baskets that support the catalyst as a fixed bed for continuous desulfurization while allowing easy filling, emptying, and regeneration. A fresh magnetic heteropolymetalate catalyst (5% MnO₂–3% Fe<sub>2</sub>O<sub>3</sub> supported on activated carbon) was synthesized, providing a high surface area, abundant active sites, and enhanced adsorption–oxidation performance. Hydrogen peroxide was employed as the oxidizing agent under mild operating conditions. Under optimal conditions (90&#xa0;°C, liquid hourly space velocity (LHSV) of 0.08&#xa0;min⁻¹, oscillation frequency of 2&#xa0;Hz, and amplitude of 12&#xa0;mm), desulfurization efficiencies of 99.64% for real diesel and 98.11% for model diesel were achieved. To establish a robust predictive framework, multiple machine-learning models, including support vector machines (SVM), gradient boosting models (GBM), and artificial neural networks (ANN), were developed and evaluated. All models exhibited high predictive accuracy (R² &gt; 0.98), with GBM outperforming the others by achieving minimal mean absolute errors, MAE (0.0258 and 0.0095) and exceptionally high R² values (0.9995 and 0.9999) for real and model diesel fuels, respectively. These results demonstrate the strong potential of the integrated reactor–catalyst–AI framework for advanced continuous desulfurization applications.</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

Integration of artificial intelligence and eco-friendly fuel production: utilizing an oscillatory basket reactor with a magnetic heteropolymetalate carbonaceous catalyst for deep desulfurization of diesel fuel

  • 期刊论文

  • 发表日期：

    2026年3月17日

查看全文
  • Jasim I. Humadi,
  • Wadood T. Mohammed

摘要

This study presents a novel integrated approach for deep desulfurization of both model and real diesel fuels through the combined development of a hybrid oscillatory-flow basket reactor, a newly synthesized magnetic heteropolymetalate carbon-based catalyst, and artificial intelligence–based predictive modeling. The proposed reactor design integrates key features of conventional basket reactors and central oscillatory-baffled reactors, forming centrally baffled baskets that support the catalyst as a fixed bed for continuous desulfurization while allowing easy filling, emptying, and regeneration. A fresh magnetic heteropolymetalate catalyst (5% MnO₂–3% Fe2O3 supported on activated carbon) was synthesized, providing a high surface area, abundant active sites, and enhanced adsorption–oxidation performance. Hydrogen peroxide was employed as the oxidizing agent under mild operating conditions. Under optimal conditions (90 °C, liquid hourly space velocity (LHSV) of 0.08 min⁻¹, oscillation frequency of 2 Hz, and amplitude of 12 mm), desulfurization efficiencies of 99.64% for real diesel and 98.11% for model diesel were achieved. To establish a robust predictive framework, multiple machine-learning models, including support vector machines (SVM), gradient boosting models (GBM), and artificial neural networks (ANN), were developed and evaluated. All models exhibited high predictive accuracy (R² > 0.98), with GBM outperforming the others by achieving minimal mean absolute errors, MAE (0.0258 and 0.0095) and exceptionally high R² values (0.9995 and 0.9999) for real and model diesel fuels, respectively. These results demonstrate the strong potential of the integrated reactor–catalyst–AI framework for advanced continuous desulfurization applications.