A dual-domain analytical strategy for holistic quality evaluation of Citrus aurantium L. flowers by maximum-absorbance HPLC-DAD fingerprinting, HS-GC-IMS, and binary evaluation.

To establish a holistic quality evaluation approach for Citrus aurantium L. flowers, this study developed an integrated dual-domain analytical strategy combining non-volatile and volatile chemical fingerprints. A maximum-absorbance HPLC-DAD chromatogram was generated by selecting the highest absorbance within 200-400 nm at each retention time. This operation, which is mathematically equivalent to the conventional photodiode-array MaxPlot, was used as a standardized input for fingerprint alignment and batch-to-batch comparison, thereby reducing the bias associated with the selection of a single detection wavelength. The resulting non-volatile fingerprints and HS-GC-IMS volatile profiles were evaluated using the Binary Evaluation System (BES), which summarizes fingerprint-shape consistency and overall abundance variation through the qualitative similarity index (Sm) and quantitative similarity index (Pm), respectively. Analysis of 24 batches from six geographical origins revealed high qualitative consistency (Sm > 0.98) but pronounced variation in overall chemical abundance (Pm, 88.9%-109.0%). Head-to-head comparison showed that Sm was highly concordant with pharmacopoeial fingerprint similarity ((r = 0.941), (P < 0.001)), whereas Pm was strongly associated with the total common-peak area ((r = 0.985), (P < 0.001)), indicating that BES provides abundance-related information complementary to conventional shape-based similarity evaluation. The HPLC-derived and GC-IMS-derived Pm values exhibited asynchronous variation and a moderate negative correlation trend ((r = -0.389), (P = 0.061)), supporting the complementary value of dual-domain profiling. DPPH radical-scavenging capacity was used as a chemical response variable for exploratory partial least-squares modeling. Neohesperidin, naringin, and several non-assigned chromatographic peaks were identified as fingerprint variables statistically associated with variation in the DPPH response. Overall, this dual-domain strategy provides a practical and interpretable framework for characterizing complementary non-volatile and volatile quality attributes in complex plant-derived materials.