Scholarlink

<p>Identification of the type of cooking oils used in processed foods is essential for food safety, quality control, and regulatory compliance, yet routine analysis remains constrained by solvent-intensive extraction procedures and limited applicability to complex food matrices. This study presents a proof-of-concept Raman-chemometric framework for oil-type differentiation in controlled samples using minimal, solvent-free sampling technique. Potato chips were selected as a representative fried food matrix, and five commonly used edible oils (sunflower, soybean, groundnut, palm, and vanaspati) were analyzed in both oil form and corresponding chips matrices to enable systematic cross-matrix evaluation. A one-step low‑lint, laboratory‑grade cellulose wipe-based blotting method was employed for rapid oil recovery, followed by Raman spectroscopic analysis. Fatty-acid-associated Raman peaks were identified and systematically combined into chemically interpretable inter-peak intensity ratios reflecting variations in saturation and ester content. A statistically grounded two-stage marker selection workflow integrating Random Forest importance ranking with non-parametric Kruskal–Wallis testing was applied, followed by one-way analysis of variance and coefficient of variation analysis. Five robust ratiometric markers (I<sub>1652/1742</sub>, I<sub>1652/1434</sub>, I<sub>1742/1259</sub>, I<sub>1652/1259</sub>, and I<sub>1296/1434</sub>) showed strong association with saturated versus unsaturated fatty acid profiles. Multivariate analysis based on these markers revealed pronounced separation among oil types (91% explained variance; <i>F</i>-values up to 22,629; <i>p</i> &lt; 0.001) and statistically significant discrimination within chips matrices (77% explained variance), despite attenuation effects from the food matrix. Multivariate analysis of variance confirmed robust separation (Wilks’ Λ = 0.0307; <i>p</i> &lt; 0.0001). Overall, this framework establishes the analytical foundation and marker identification in controlled fried food samples.</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

Raman-Chemometric Framework for Rapid Differentiation of Edible Oils in Processed Foods Using a One-Step Sampling Technique—Proof-of-Concept Study

  • 期刊论文

  • 发表日期：

    2026年6月2日

查看全文
  • Amrita Shaw,
  • Chandrasekar S. N.,
  • Sai Muthukumar V.,
  • Deepak L. N. Kallepalli,
  • Bhanu Prakash Rachaiah,
  • Jhinuk Gupta

摘要

Identification of the type of cooking oils used in processed foods is essential for food safety, quality control, and regulatory compliance, yet routine analysis remains constrained by solvent-intensive extraction procedures and limited applicability to complex food matrices. This study presents a proof-of-concept Raman-chemometric framework for oil-type differentiation in controlled samples using minimal, solvent-free sampling technique. Potato chips were selected as a representative fried food matrix, and five commonly used edible oils (sunflower, soybean, groundnut, palm, and vanaspati) were analyzed in both oil form and corresponding chips matrices to enable systematic cross-matrix evaluation. A one-step low‑lint, laboratory‑grade cellulose wipe-based blotting method was employed for rapid oil recovery, followed by Raman spectroscopic analysis. Fatty-acid-associated Raman peaks were identified and systematically combined into chemically interpretable inter-peak intensity ratios reflecting variations in saturation and ester content. A statistically grounded two-stage marker selection workflow integrating Random Forest importance ranking with non-parametric Kruskal–Wallis testing was applied, followed by one-way analysis of variance and coefficient of variation analysis. Five robust ratiometric markers (I1652/1742, I1652/1434, I1742/1259, I1652/1259, and I1296/1434) showed strong association with saturated versus unsaturated fatty acid profiles. Multivariate analysis based on these markers revealed pronounced separation among oil types (91% explained variance; F-values up to 22,629; p < 0.001) and statistically significant discrimination within chips matrices (77% explained variance), despite attenuation effects from the food matrix. Multivariate analysis of variance confirmed robust separation (Wilks’ Λ = 0.0307; p < 0.0001). Overall, this framework establishes the analytical foundation and marker identification in controlled fried food samples.

Graphical Abstract