Joan Masanet Martí, Antonio Belda-Antolí, Daniel López Rodríguez, Jorge Jordán-Núñez, Álvaro-Francisco Morote, Bàrbara Micó-Vicent
Replacing synthetic colorants with natural alternatives requires source-specific extraction strategies because pigment classes differ in polarity, pH response and thermal stability. This first-pass screening focused on six plant and agro-food matrices: red cabbage, Iris germanica petals, dragon fruit, orange peel, spent coffee grounds and pine bark. Each matrix was subjected to twelve extraction treatments combining water, ethanol, acetone 30%, acid/base modifiers, boiling water, freezing and ultrasound assistance. Extracts were evaluated by visible transmittance spectroscopy from 360 to 740 nm, operationally transformed to apparent optical attenuation according to Aapp (λ) = -log10 [%T (λ)/100] only where measured %T exceeded 0.01; this signal may include both absorption and light scattering, and converted to CIELAB coordinates under D65/10° conditions. The combined spectral and colorimetric approach separated the extracts into red-purple and yellow-brown groups. Red cabbage and Iris showed anthocyanin-like behaviour, dragon fruit showed a betalain-compatible magenta response, orange peel showed carotenoid-compatible short-wavelength attenuation, and coffee grounds and pine bark generated broad brown-yellow profiles. Pigment-family assignments remain putative because chromatographic identification was not performed. This study is explicitly intended as a proof-of-concept descriptive mapping step rather than as a statistically validated extraction optimisation. Because every matrix-treatment combination was prepared once, all comparisons in this article describe the observed screening preparations only; they do not estimate reproducible treatment effects, confidence intervals or statistical significance. The previous equal-weight low-L*/high-C*ab intensity score has been removed from the revised analysis. No extraction yield, recovery percentage, mass balance, purity, pigment concentration or application performance was measured; accordingly, this study describes the optical appearance of the obtained extracts and does not quantify colorant recovery. Visible transmittance, apparent optical attenuation and CIELAB coordinates support only literature-informed class-level hypotheses and cannot identify a pigment family or individual compound.