Prediction of the flash point of fragrances by raman spectroscopy and multivariate calibration

In this work, we investigated the use of Raman spectroscopy for rapid, nondestructive, and low‐cost determination of the flash point in fragrances. For this purpose, multivariate calibration based on partial least squares regression was applied to correlate the fragrances Raman spectra with the respective flash point estimated by the ASTM D7094‐04 method. To perform this study, 32 fragrances having distinct odor profiles were chosen (referred to as “fragrance type” in the following), corresponding to a total of 81 batches produced in both factory and laboratory. For the samples in the prediction set, the correlation coefficient for the linear adjust between the flash point estimated by proposed methodology and the ASTM method was 0.9901, and the largest absolute error was 5.2°C. The proposed methodology presented an inverse of the analytical sensitivity of 2.1°C, that is, the minimum flash point difference which is discernible by the model. The results demonstrated that the proposed method is comparable to the standard method for analyzing the flash points of fragrances. Raman spectroscopy was applied for rapid, nondestructive and low‐cost determination of the flash point in fragrances.