Scholarlink

<p>This study optimized the sonication-assisted hydrodistillation (SAHD) process for extracting essential oil (EO) from <i>Cinnamomum tamala</i> leaves, aiming to maximize yield and antioxidant activity. Both response surface methodology (RSM) and artificial neural network (ANN) models were used to predict extraction performance. The interpretability of the ANN model was enhanced using a neural interpretation diagram (NID), Olden’s algorithm, and sensitivity analysis, and it demonstrated higher accuracy and generalization than RSM. Under optimized conditions, the EO yield reached 1.67 ± 0.13%, with strong antioxidant activity indicated by a total phenolic content (TPC) of 79.24 ± 0.82&#xa0;mg GAE/g and 81.54 ± 0.88% inhibition of the 2,2-diphenyl-1-picrylhydrazyl radical (DPPH). Residual analysis showed that both models satisfied key regression assumptions, including normality, independence, homoscedasticity, and lack of bias. Gas chromatography–mass spectrometry (GC–MS) analysis identified linalool (47.37%), eugenol (18.34%), and cinnamaldehyde (16.45%) as key constituents. Physicochemical characterization verified EO quality and stability. The integrated modeling approach provides a robust framework for enhancing EO extraction efficiency.</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

Optimizing sonication-assisted hydrodistillation of Cinnamomum tamala essential oil using response surface methodology and artificial neural network modeling

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月18日

查看全文
  • Parvej Hasan Jon,
  • Jahid Hasan Shourove,
  • Md. Kashem Ali,
  • Oliur Rahman,
  • Mostak Uddin Thakur,
  • G. M. Rabiul Islam

摘要

This study optimized the sonication-assisted hydrodistillation (SAHD) process for extracting essential oil (EO) from Cinnamomum tamala leaves, aiming to maximize yield and antioxidant activity. Both response surface methodology (RSM) and artificial neural network (ANN) models were used to predict extraction performance. The interpretability of the ANN model was enhanced using a neural interpretation diagram (NID), Olden’s algorithm, and sensitivity analysis, and it demonstrated higher accuracy and generalization than RSM. Under optimized conditions, the EO yield reached 1.67 ± 0.13%, with strong antioxidant activity indicated by a total phenolic content (TPC) of 79.24 ± 0.82 mg GAE/g and 81.54 ± 0.88% inhibition of the 2,2-diphenyl-1-picrylhydrazyl radical (DPPH). Residual analysis showed that both models satisfied key regression assumptions, including normality, independence, homoscedasticity, and lack of bias. Gas chromatography–mass spectrometry (GC–MS) analysis identified linalool (47.37%), eugenol (18.34%), and cinnamaldehyde (16.45%) as key constituents. Physicochemical characterization verified EO quality and stability. The integrated modeling approach provides a robust framework for enhancing EO extraction efficiency.