Scholarlink

<p>Accurate identification and differentiation of orange cultivars are essential for food quality control, market segmentation and product authentication. Conventional approaches rely on multiple physicochemical determinations and sensory evaluation, resulting in labor-intensive, reagent-consuming analytical workflows. In this study, a lab-made potentiometric electronic tongue (E-tongue) is evaluated as a rapid and reagent-free alternative for discriminating freshly squeezed orange juices according to cultivar. Five commercially relevant orange cultivars (Baía, Dalmau, Lane Late, Navelate and Salustiana) were characterized using fruit morphological descriptors, juice physicochemical parameters (soluble solids: 9.1–12.9 °Brix, titratable acidity: 0.7–0.8&#xa0;g citric acid·100&#xa0;g<sup>-1</sup>, total solids: up to 13.1%) and consumer sensory evaluation, and compared with direct potentiometric fingerprints acquired without reagents, solvents or sample pretreatment. Multivariate analysis showed that a linear discriminant model based on nine selected morphological and physicochemical variables, selected by the simulated annealing algorithm, achieved an average correct classification of 89% under repeated 4-fold cross-validation, whereas the E-tongue-based model, relying on ten selected potentiometric sensors, provided improved robustness with an average sensitivity of 94% under the same validation scheme. The potentiometric analysis required a single 5&#xa0;min measurement per sample and generated negligible waste. These findings demonstrate that potentiometric fingerprinting can serve as an effective alternative or complementary approach to conventional multi-parameter analytical workflows for orange cultivar identification, providing robust discriminatory capability while substantially reducing analytical complexity, analysis time, and resource consumption. The proposed approach represents a practical tool for food quality assessment and routine screening applications where rapid and reproducible cultivar discrimination is required.</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

Orange cultivars identification: electronic tongue versus conventional physicochemical methods

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月28日

查看全文
  • Valéria Serra,
  • Ana C. A. Veloso,
  • António M. Peres,
  • Marta H. F. Henriques,
  • Ana S. Fajardo

摘要

Accurate identification and differentiation of orange cultivars are essential for food quality control, market segmentation and product authentication. Conventional approaches rely on multiple physicochemical determinations and sensory evaluation, resulting in labor-intensive, reagent-consuming analytical workflows. In this study, a lab-made potentiometric electronic tongue (E-tongue) is evaluated as a rapid and reagent-free alternative for discriminating freshly squeezed orange juices according to cultivar. Five commercially relevant orange cultivars (Baía, Dalmau, Lane Late, Navelate and Salustiana) were characterized using fruit morphological descriptors, juice physicochemical parameters (soluble solids: 9.1–12.9 °Brix, titratable acidity: 0.7–0.8 g citric acid·100 g-1, total solids: up to 13.1%) and consumer sensory evaluation, and compared with direct potentiometric fingerprints acquired without reagents, solvents or sample pretreatment. Multivariate analysis showed that a linear discriminant model based on nine selected morphological and physicochemical variables, selected by the simulated annealing algorithm, achieved an average correct classification of 89% under repeated 4-fold cross-validation, whereas the E-tongue-based model, relying on ten selected potentiometric sensors, provided improved robustness with an average sensitivity of 94% under the same validation scheme. The potentiometric analysis required a single 5 min measurement per sample and generated negligible waste. These findings demonstrate that potentiometric fingerprinting can serve as an effective alternative or complementary approach to conventional multi-parameter analytical workflows for orange cultivar identification, providing robust discriminatory capability while substantially reducing analytical complexity, analysis time, and resource consumption. The proposed approach represents a practical tool for food quality assessment and routine screening applications where rapid and reproducible cultivar discrimination is required.