Non-destructive biochemical profiling of goldenberry ripening using confocal micro Raman spectroscopy coupled with...

Area of Science:

• * Agricultural Science
• * Analytical Chemistry
• * Food Science

Background:

• * Accurate fruit ripening monitoring is vital for quality and preventing post-harvest economic losses.
• * Traditional goldenberry ripening assessment methods are subjective and variable.
• * Goldenberries are nutritious, rich in vitamins, minerals, and carotenoids.

Purpose of the Study:

• * To develop rapid, non-destructive techniques for assessing goldenberry ripening stages.
• * To track biochemical changes (carotenoids, chlorophyll) in exocarp and mesocarp during ripening.
• * To evaluate the effectiveness of machine learning models in classifying ripening stages.

Main Methods:

• * Confocal micro-Raman and UV-Vis spectroscopy were employed to analyze fruit tissues.
• * Five ripening stages were studied: pre-mature, mature, pre-ripe, ripe, and post-ripe.
• * Various machine learning algorithms (k-means, SVM, Random Forest, PLS regression) were utilized.

Main Results:

• * Raman spectroscopy revealed changes in carotenoid and chlorophyll content correlating with ripening stages.
• * UV-Vis spectroscopy confirmed increasing carotenoids and decreasing chlorophyll during ripening.
• * Machine learning models, particularly k-means and SVM, showed high accuracy (>0.97) in classifying ripening stages.
• * Partial Least Squares regression demonstrated strong predictive power for carotenoid content (R² up to 0.93).

Conclusions:

• * The integration of Raman spectroscopy with machine learning provides a reliable framework for objective fruit ripening assessment.
• * This non-destructive approach offers practical applications in goldenberry quality control and post-harvest management.
• * The study highlights the potential for advanced spectroscopic and computational methods in agricultural product assessment.