Laboratório de Espectrometria, Sensores e Biossensores

The determination of the sun protection factor (SPF) remains a complex and resource-intensive task, motivating the development of alternative approaches. In this study, a chemometric-assisted strategy for SPF classification based on in vitro UV spectral data is proposed. UV absorbance spectra were obtained according to ISO 24443:2012 and analyzed using self-organizing maps (SOM) and principal component analysis (PCA). Classification models were developed using partialleast squares discriminant analysis (PLS-DA). The exploratory analysis indicated that spectral variations, mainly associated with absorbance intensity in the 290–340 nm region, are relevant for distinguishing SPF classes (15, 30, and 50). The PLS-DA models achieved classification rates of 93% for the training set and 100% for the external validation set, with sensitivities above 87%. These results demonstrate that the proposed approach can be used as a rapid and cost-effective tool for SPF classification. Although not intended to replace standardized in vivo methods, this strategy may support sunscreen development and preliminary evaluation. Further studies are required to improve model robustness and expand its applicability.