Influence of drying temperature on the retention of bioactive compounds and antioxidant potential of roselle (Hibiscus sabdariffa L.): An RSM-based optimization study

Roselle ( Hibiscus sabdariffa L .) calyces are a rich source of anthocyanins, phenolic acids, flavonoids, and organic acids, rendering them a promising raw material for functional food applications. The stability of these bioactive compounds during hot-air drying and subsequent storage remains an important determinant of product quality. This study investigated the combined effects of drying temperature (60–80 °C) and storage duration (0–90 days) on total anthocyanin content (TAC), total phenolic content (TPC), total flavonoid content (TFC), and DPPH free radical scavenging activity of dried roselle calyces using Response Surface Methodology (RSM) with a full-factorial design (3 drying temperatures × 7 storage durations, 21 runs). Second-order polynomial models were fitted to all responses (R² > 0.95; Adequate Precision > 24), and ANOVA confirmed their statistical significance (p < 0.0001). Anthocyanin content declined monotonically with increasing storage time, driven primarily by oxidative and hydrolytic degradation pathways. Total phenolic and flavonoid contents exhibited nonlinear responses to drying temperature, with U-shaped curves showing higher values at 60 and 80 °C than at 70 °C, plausibly linked to differential enzyme inactivation and cell-wall disruption, and declined progressively with storage duration owing to oxidative polymerisation. DPPH radical scavenging activity was predominantly governed by storage duration and declined progressively over time, from 92.30% to 79.28% at 60 °C and from 90.10% to 80.03% at 80 °C across the 90-day storage period, at a proportionally slower rate than anthocyanin loss, possibly reflecting a partial compensatory contribution from secondary antioxidant species formed during storage. Numerical multi-response optimization of drying temperature alone identified 77.87 °C as the condition maximising fresh-dried quality, since all four responses were highest at zero storage and declined thereafter; storage duration was therefore treated as a shelf-life variable rather than a controllable process factor, and a representative 42.4-day shelf-life point was evaluated to illustrate predicted quality retention at that stage, giving TAC (13.08 mg CGE/100 g dw), TPC (14.94 mg GAE/100 g dw), TFC (13.38 mg QE/100 g dw), and DPPH activity (85.25%) with an overall desirability index of 0.95. Experimental validation under these conditions yielded deviations of less than 5% from predicted values, confirming model reliability. The findings provide actionable guidelines for the design of quality-preserving drying and storage protocols for roselle-based nutraceutical and functional food products, while highlighting the need for complementary energy- and cost-related data to fully establish their sustainability credentials.