RESUMO
Two new procedures, based on infrared spectroscopy in the near infrared (NIR) and mid infrared (MIR), have been developed for the authentication of tea tree oil (TTO) commercial samples. Infrared measurements were made on untreated samples by transmission NIR and attenuated total reflectance (ATR) followed by partial least square discriminant analysis (PLS-DA). These methods offer a fast and low cost alternative to enantiomeric two-dimensional gas chromatography coupled to mass detection usually employed to discriminate between authentic and non-authentic samples. In these studies, a set of 267 samples, including authentic and non-authentic labelled tea tree oil samples, were used to build the models based on the wavenumber range, data pre-processing and latent variables number selection. Infrared methods can be discriminant for authentic and non-authentic TTO samples with a 98% certainty for both ATR and NIR methodologies, employing 92 and 142 external samples respectively. Developed PLS-DA infrared based methodologies and the reference methodology have been evaluated and compared from a Green Analytical point of view.
