Scholarlink

<p>Naproxen (NPX) is primarily used to treat fever, inflammation, and pain. However, its extensive use and resistance to degradation in wastewater treatment processes lead to its release into water bodies, necessitating the development of efficient removal approaches. Chitosan-based functional adsorbents were synthesized in this study by grafting <i>β</i>-cyclodextrin and incorporating calcium ions. The structural and compositional features of the adsorbents were characterized using a range of analytical methods. Adsorption experiments were performed to evaluate relevant adsorption models and understand the adsorption mechanisms. The experimental maximum adsorption capacity was found to be 314.6&#xa0;mg/g, and the adsorption behavior was characterized as exothermic. The adsorption mechanism involved hydrogen bonding, electrostatic interactions, and metal cation–π interactions. Furthermore, a Box–Behnken design was employed to generate experimental data for deep learning modeling, incorporating four input variables: pH, adsorbent dosage, initial NPX concentration, and temperature. The NPX removal rate was used as the output variable. Artificial neural network models—radial basis function network (RBFN) and multi-layer perceptron (MLP)—were developed to optimize NPX adsorption efficiency and determine the most influential input variable. Model analysis with additional experiments revealed that the MLP model outperformed the RBFN model, achieving a higher coefficient of determination. Relative importance analysis identified initial NPX concentration (100%) as the most influential factor, followed by adsorbent dosage (88.5%), temperature (83.2%), and pH (79.4%).</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

Adsorption experiments and deep-learning modeling for characterization of naproxen adsorption on chitosan-based functional adsorbents in aqueous solutions

  • 期刊论文

  • 发表日期：

    2025年11月26日

查看全文
  • S.-C. Lee,
  • S.-B. Kim

摘要

Naproxen (NPX) is primarily used to treat fever, inflammation, and pain. However, its extensive use and resistance to degradation in wastewater treatment processes lead to its release into water bodies, necessitating the development of efficient removal approaches. Chitosan-based functional adsorbents were synthesized in this study by grafting β-cyclodextrin and incorporating calcium ions. The structural and compositional features of the adsorbents were characterized using a range of analytical methods. Adsorption experiments were performed to evaluate relevant adsorption models and understand the adsorption mechanisms. The experimental maximum adsorption capacity was found to be 314.6 mg/g, and the adsorption behavior was characterized as exothermic. The adsorption mechanism involved hydrogen bonding, electrostatic interactions, and metal cation–π interactions. Furthermore, a Box–Behnken design was employed to generate experimental data for deep learning modeling, incorporating four input variables: pH, adsorbent dosage, initial NPX concentration, and temperature. The NPX removal rate was used as the output variable. Artificial neural network models—radial basis function network (RBFN) and multi-layer perceptron (MLP)—were developed to optimize NPX adsorption efficiency and determine the most influential input variable. Model analysis with additional experiments revealed that the MLP model outperformed the RBFN model, achieving a higher coefficient of determination. Relative importance analysis identified initial NPX concentration (100%) as the most influential factor, followed by adsorbent dosage (88.5%), temperature (83.2%), and pH (79.4%).