Silver nanocube-enabled surface-enhanced Raman scattering platform for predictive quantification and classification modeling of white colony-forming yeasts in kimchi

White colony-forming yeasts (WCFYs), including Candida sake and Kazachstania servazzii, frequently appear during kimchi fermentation, yet rapid and accurate detection methods remain limited. In this study, we developed a highly sensitive surface-enhanced Raman scattering (SERS) platform by immobilizing silver nanocubes (AgNCs) onto a Si wafer using a 3-aminopropyltriethoxysilane (APTES)-mediated method. The resulting AgNC_Si SERS substrate exhibited dense nanocube packing (average edge length: 107 nm; interparticle gap: 4.7 nm), as confirmed via scanning electron microscopy and energy-dispersive X-ray spectroscopy. Finite-difference time-domain simulations revealed intense electric-field enhancement at cube edges and junctions. The SERS signals of WCFYs exhibited linear increases in the intensities of peaks at 1562 and 2876 cm⁻¹ with increasing yeast concentrations, achieving limits of detection as low as 1.00–1.02 CFU/mL. In spiked kimchi samples, K. servazzii was quantitatively predicted with recovery rates of 94.48 %–101.72 % using the 2876-cm⁻¹ peak. Furthermore, a principal component analysis (PCA) and linear discriminant analysis (LDA) classification model achieved 90 % accuracy, effectively discriminating C. sake, K. servazzii, and the kimchi matrix. These findings demonstrated the potential of the AgNC_Si SERS platform for rapid, onsite detection and monitoring of WCFYs in complex fermented-food systems.