Scholarlink

<p>Despite having numerous dye-based treatment studies, there still exists a key challenge to determine the optimal process conditions for effective dye removal using experimental modelling. RSM and ANN are two of the influential modeling techniques in this regard. However, there are considerable knowledge gaps in determining their strength and weakness during prediction. Thus, this study is an attempt to understand the prediction abilities of both techniques in modeling the appropriate dye-removal conditions using GSH@Fe<sub>3</sub>O<sub>4</sub> core-shell nanoparticles in light of accuracy and resource management. Herein, GSH@Fe<sub>3</sub>O<sub>4</sub> nanoparticles were employed for the removal of crystal violet (CV) dye from an aqueous system through experimental modeling. The nanoparticles were characterized for their surface physico-chemical properties before and after dye removal. RSM-CCD model designed 50 experimental runs based on 5 input variables, i.e. adsorbent-dose, pH, reaction time, temperature, and initial dye-concentration to provide the output response, CV-removal %. The highest CV removal, i.e. 91.1% was observed when the GSH@Fe<sub>3</sub>O<sub>4</sub> MNPs dose was 1.88 mg, pH-11, initial CV dye concentration-100 mg/L, reaction time-50 minutes at room temperature. The quadratic-regression model of RSM-CCD design was found adequate in describing the CV dye adsorption by ANOVA-analysis with an R<sup>2</sup> value of 0.99 and predicted R<sup>2</sup> value of 0.98. A feed-forward backdrop network of 5:14:1 ANN architecture with Levenberg-Marquardt (LM) backpropagation algorithm was employed for training datasets. The ANN model was trained using the data obtained from the RSM-CCD model under the optimized conditions. The final modeling results were in best agreement with the experimental data with an R<sup>2</sup> -0.98 along with high <i>R</i> values of 0.98, 0.99, and 0.99 for training, validation, and test datasets respectively. Further, the mass analysis of residual CV samples in LC-MS indicated partial degradation of the dye molecules. Overall, the study’s proposition of integrating RSM and ANN models to optimize and process control CV-dye removal with high accuracy has been found to be highly promising and could contribute to sustainable environmental engineering practices.</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

Prediction and optimization of cationic textile dye adsorption using glutathione functionalized magnetic core-shell nanoparticles: An integrated RSM-ANN modelling approach

  • 期刊论文

  • 发表日期：

    2025年11月18日

查看全文
  • Monalisha Behera,
  • Nisha Kumari,
  • Prashant Kumar Jaiswal,
  • Abideen Idowu Adeogun,
  • Prashant Basavaraj Bhagawati,
  • Ritu Singh

摘要

Despite having numerous dye-based treatment studies, there still exists a key challenge to determine the optimal process conditions for effective dye removal using experimental modelling. RSM and ANN are two of the influential modeling techniques in this regard. However, there are considerable knowledge gaps in determining their strength and weakness during prediction. Thus, this study is an attempt to understand the prediction abilities of both techniques in modeling the appropriate dye-removal conditions using GSH@Fe3O4 core-shell nanoparticles in light of accuracy and resource management. Herein, GSH@Fe3O4 nanoparticles were employed for the removal of crystal violet (CV) dye from an aqueous system through experimental modeling. The nanoparticles were characterized for their surface physico-chemical properties before and after dye removal. RSM-CCD model designed 50 experimental runs based on 5 input variables, i.e. adsorbent-dose, pH, reaction time, temperature, and initial dye-concentration to provide the output response, CV-removal %. The highest CV removal, i.e. 91.1% was observed when the GSH@Fe3O4 MNPs dose was 1.88 mg, pH-11, initial CV dye concentration-100 mg/L, reaction time-50 minutes at room temperature. The quadratic-regression model of RSM-CCD design was found adequate in describing the CV dye adsorption by ANOVA-analysis with an R2 value of 0.99 and predicted R2 value of 0.98. A feed-forward backdrop network of 5:14:1 ANN architecture with Levenberg-Marquardt (LM) backpropagation algorithm was employed for training datasets. The ANN model was trained using the data obtained from the RSM-CCD model under the optimized conditions. The final modeling results were in best agreement with the experimental data with an R2 -0.98 along with high R values of 0.98, 0.99, and 0.99 for training, validation, and test datasets respectively. Further, the mass analysis of residual CV samples in LC-MS indicated partial degradation of the dye molecules. Overall, the study’s proposition of integrating RSM and ANN models to optimize and process control CV-dye removal with high accuracy has been found to be highly promising and could contribute to sustainable environmental engineering practices.