Abstract
White croaker surimi, an important raw material of traditional kamaboko industry, commonly has four grades, A, AA, FA, and SA which have distinct quality but are tough to be identified and discriminated rapidly and holistically by conventional methods. A Tri-step infrared spectroscopy, Fourier transform infrared spectroscopy (FT-IR) integrated with second derivative infrared (SD-IR) spectroscopy, and two-dimensional correlation infrared spectroscopy (2DCOS-IR), combined with statistical pattern recognition, was employed to rapidly characterize and discriminate four grades of white croaker surimi (A, AA, FA, and SA). The four surimis had similar IR, and SD-IR macro-fingerprints (similarity >0.9) especially for the absorption bands of amide groups due to proteins were the main compositions. However, relative contents of lipids in all surimis could be directly observed. Compared to the other three surimis, SA had a middle strong characteristic peak of lipids at 1745 cm−1, indicating that SA had the highest content of lipids. The sequence of lipid contents in the four surimis was SA > FA > AA > A, which was further verified by SD-IR spectra. Moreover, evident differences were disclosed visually in 2DCOS-IR spectra of 1680–1610 cm−1, indicating that the four surimis have fine differences in protein secondary structures. Furthermore, 76 parallel samples (18 A, 19 AA, 19 FA, and 20 SA) were objectively classified by soft independent modeling of class analogy (SIMCA) based on principal component regression (PCR). It has been demonstrated that the Tri-step infrared spectroscopy combined with cluster analysis could be a scientific and powerful tool for rapid discrimination and identification of different grades of surimi in a holistic manner.
