Scholarlink

<p>Phthalate esters, commonly used as plasticizers, can leach from materials that come into contact with food into edible oils, posing significant risks to food safety and human health. This study presents a simple fluorescence-based chemometric workflow with low hands-on time and short measurement time after overnight equilibration for discriminating seven phthalate esters in extra virgin olive oil (EVOO) using a BODIPY-Me/γ-cyclodextrin supramolecular sensing system. Commercial EVOO samples were spiked with individual phthalate standards (dibutyl phthalate-DBP, dicyclohexyl phthalate-DCHP, di(2-ethylhexyl) phthalate-DEHP, diethyl phthalate-DEP, diisodecyl phthalate-DIDP, dimethyl phthalate-DMP, and di-n-octyl phthalate-DnOP), extracted with isooctane, and balanced with an aqueous γ-CD/phosphate-buffered saline (PBS, pH 7.4) solution containing BODIPY-Me. After a 24-hour equilibration period, fluorescence emission spectra (470–700&#xa0;nm, λex = 460&#xa0;nm) were captured from the aqueous phase. The dataset included 72 samples (Blank + 7 phthalates, 9 replicates each) with 231 wavelengths per spectrum. Peak fluorescence intensities and F/F₀ ratios compared to the Blank exhibited significant class-dependent changes Principal component analysis showed that for the full 470–700&#xa0;nm region, PC1 accounted for 91.5% of the spectral variance. In the optimal 480–600&#xa0;nm region, PCA showed a much more compact variance structure, with PC1 explaining 97.42% of the variance and clear class separation. The optimal spectral range (480–600&#xa0;nm) provided outstanding classification results: PCA-LDA achieved 95.7% accuracy (5-fold cross-validation), while Logistic Regression attained 97.2% accuracy across all eight classes. Additional classifiers (SVM-RBF and Random Forest) were assessed, with linear models showing superior performance. This efficient method offers a reliable screening tool for detecting phthalate contamination in EVOO, which can be easily adaptable to other edible oil matrices for food composition and safety assessments.</p> Graphical Abstract <p></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

Fluorescence-Based Chemometric Discrimination of 7 Phthalates Esters in Extra Virgin Olive Oil Using BODIPY-Me and γ-Cyclodextrin

  • 期刊论文

  • 发表日期：

    2026年7月28日

查看全文
  • Kunal Shiv,
  • Ankit,
  • Sachin Kumar,
  • Anupam Singh,
  • Lal Bahadur Prasad,
  • Manoj Kumar Bharty

摘要

Phthalate esters, commonly used as plasticizers, can leach from materials that come into contact with food into edible oils, posing significant risks to food safety and human health. This study presents a simple fluorescence-based chemometric workflow with low hands-on time and short measurement time after overnight equilibration for discriminating seven phthalate esters in extra virgin olive oil (EVOO) using a BODIPY-Me/γ-cyclodextrin supramolecular sensing system. Commercial EVOO samples were spiked with individual phthalate standards (dibutyl phthalate-DBP, dicyclohexyl phthalate-DCHP, di(2-ethylhexyl) phthalate-DEHP, diethyl phthalate-DEP, diisodecyl phthalate-DIDP, dimethyl phthalate-DMP, and di-n-octyl phthalate-DnOP), extracted with isooctane, and balanced with an aqueous γ-CD/phosphate-buffered saline (PBS, pH 7.4) solution containing BODIPY-Me. After a 24-hour equilibration period, fluorescence emission spectra (470–700 nm, λex = 460 nm) were captured from the aqueous phase. The dataset included 72 samples (Blank + 7 phthalates, 9 replicates each) with 231 wavelengths per spectrum. Peak fluorescence intensities and F/F₀ ratios compared to the Blank exhibited significant class-dependent changes Principal component analysis showed that for the full 470–700 nm region, PC1 accounted for 91.5% of the spectral variance. In the optimal 480–600 nm region, PCA showed a much more compact variance structure, with PC1 explaining 97.42% of the variance and clear class separation. The optimal spectral range (480–600 nm) provided outstanding classification results: PCA-LDA achieved 95.7% accuracy (5-fold cross-validation), while Logistic Regression attained 97.2% accuracy across all eight classes. Additional classifiers (SVM-RBF and Random Forest) were assessed, with linear models showing superior performance. This efficient method offers a reliable screening tool for detecting phthalate contamination in EVOO, which can be easily adaptable to other edible oil matrices for food composition and safety assessments.

Graphical Abstract