Simultaneous quantification of the boar-taint compounds skatole and androstenone by surface-enhanced Raman scattering (SERS) and multivariate data analysis

This study investigates the feasibility of using surface-enhanced Raman scattering (SERS) for the quantification of absolute levels of the boar-taint compounds skatole and androstenone in porcine fat. By investigation of different types of nanoparticles, pH and aggregating agents, an optimized environment that promotes SERS of the analytes was developed and tested with different multivariate spectral pre-processing techniques, and this was combined with variable selection on a series of analytical standards. The resulting method exhibited prediction errors (root mean square error of cross validation, RMSECV) of 2.4 × 10⁻⁶ M skatole and 1.2 × 10⁻⁷ M androstenone, with a limit of detection corresponding to approximately 2.1 × 10⁻¹¹ M for skatole and approximately 1.8 × 10⁻¹⁰ for androstenone. The method was subsequently tested on porcine fat extract, leading to prediction errors (RMSECV) of 0.17 μg/g for skatole and 1.5 μg/g for androstenone. It is clear that this optimized SERS method, when combined with multivariate analysis, shows great potential for optimization into an on-line application, which will be the first of its kind, and opens up possibilities for simultaneous detection of other meat-quality metabolites or pathogen markers.