Effect of Molecular Weight of Poly(vinyl alcohol) on Quercetin Release from Poly(vinyl alcohol) Films

Antioxidant active packaging delays lipid oxidation by transferring a bioactive compound from the polymer to the food, so the rate of that transfer is the property that matters. Poly(vinyl alcohol) (PVA) is a biodegradable, non-toxic film-former, but how its molecular weight governs polyphenol release has not been resolved. Films were cast from PVA of low (31–50 kDa), medium (85–124 kDa) and high (146–186 kDa) molecular weight in four formulations: unloaded (F1), quercetin pre-dissolved in ethanol at 0.5 (F2) and 1.0 g L−1 (F3), and quercetin added as a powder at 1.0 g L−1 (F4). The FT-IR of all four formulations showed the crystalline quercetin band set, led by the aromatic C=C mode at 1520 cm−1, only in F4; F3 carried the same loading and showed none of it, so the method of incorporation rather than the loading decides whether a crystalline quercetin phase forms. The X-ray diffraction of all four formulations confirmed a separate crystalline phase in F4 alone: the mean height of the four diagnostic quercetin reflections was 0.013–0.033 relative to the PVA (101) reflection in F1, F2 and F3 and 0.931 in F4, with the 12.54° reflection reaching 2.59–2.89 times the polymer reflection. Atomic force microscopy found both 1.0 g L−1 formulations rougher than the control at every molecular weight, regardless of whether a crystalline phase was present, so surface relief at this loading did not report crystallinity; the two differed in roughness from each other as much as from the control, so the method of incorporation affect topography as well. The antioxidant assays add the functional consequence of the structural difference: in the fatty-food simulant, the only medium in which quercetin was freely soluble, the dispersed form yielded 1.8 times the DPPH activity and 2.1 times the extractable phenolic content of the crystalline form at the same loading, while in the two aqueous simulants, they were indistinguishable. Quercetin release in PBS over 12 days, analysed as a mixed-effects model with a random intercept for each of the 26 specimens, was governed by a molecular weight × formulation interaction (χ2=31.8 on 4 d.f., p<0.0001) with neither main effect significant alone. F3 and F4 released statistically identical total amounts (28.4 and 29.2 μmol g−1, p=0.72) but in opposite molecular-weight orders (F(2,11)=6.61, p=0.013): the non-crystalline form released 15.9 μmol g−1 more from the low-molecular-weight matrix, the crystalline form 13.8 μmol g−1 more from the high-molecular-weight one. Across ABTS, DPPH and Folin–Ciocalteu assays, the extractable antioxidant was governed chiefly by the simulant, decreasing as ethanol 95% > ethanol 10% > acetic acid 3%, consistent with the poor aqueous solubility of quercetin, and was highest from the low-molecular-weight films in all three assays. Surface roughness also fell with molecular weight from Sq=114.9 to 30.3 nm in the unloaded films. The practical conclusion is that quercetin loading and the method of its incorporation must be chosen together with the polymer grade, because the incorporation method determines which molecular weight releases fastest.