Scholarlink

<p>This study focuses on the optimization of the synthesis tetrazoles using the heterogeneous Fe/SBA catalyst. By immobilizing the catalyst on a solid support, this approach facilitates catalyst recovery, reduces the use of organic solvents, and allows for milder reaction conditions, enhancing both the environmental and economic sustainability of the process. The optimization is based on adjusting key parameters, including temperature (90&#xa0;°C), catalyst mass (0.05&#xa0;g), H<sub>2</sub>SO<sub>4</sub> concentration (0.24&#xa0;g), pH (6), and reaction time (120&#xa0;min), achieving a maximum yield of 98%. Additionally, neural networks were employed to model the complex interactions between these variables and provide accurate predictions, with a coefficient of determination (R<sup>2</sup>) of 0.94. The performance evaluation also relies on several metrics: Root Mean Square Error (RMSE) of 3.05, Mean Absolute Error (MAE) of 2.35, and Mean Squared Error (MSE) of 9.31. These findings demonstrate that combining artificial intelligence with rigorous experimental methods significantly enhances the efficiency of catalytic processes, paving the way for sustainable and optimized industrial 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

Modeling and parametric study of benzyl thiocyanate synthesis via heterogeneous iron catalyst using artificial neural networks

  • 期刊论文

  • 发表日期：

    2025年6月18日

查看全文
  • Attou Rekia,
  • Bailiche Zahra,
  • Beldjilali Mohammed,
  • Fekih Nadia,
  • Berrichi Amina,
  • Datoussaid Yazid

摘要

This study focuses on the optimization of the synthesis tetrazoles using the heterogeneous Fe/SBA catalyst. By immobilizing the catalyst on a solid support, this approach facilitates catalyst recovery, reduces the use of organic solvents, and allows for milder reaction conditions, enhancing both the environmental and economic sustainability of the process. The optimization is based on adjusting key parameters, including temperature (90 °C), catalyst mass (0.05 g), H2SO4 concentration (0.24 g), pH (6), and reaction time (120 min), achieving a maximum yield of 98%. Additionally, neural networks were employed to model the complex interactions between these variables and provide accurate predictions, with a coefficient of determination (R2) of 0.94. The performance evaluation also relies on several metrics: Root Mean Square Error (RMSE) of 3.05, Mean Absolute Error (MAE) of 2.35, and Mean Squared Error (MSE) of 9.31. These findings demonstrate that combining artificial intelligence with rigorous experimental methods significantly enhances the efficiency of catalytic processes, paving the way for sustainable and optimized industrial applications.