Photoacoustic spectroscopy and multivariate techniques: New applications to simplify the analysis of raw propolis

This paper investigated the application of photoacoustic spectroscopy to estimate total phenolics content in raw propolis collected in the state of Paran ' a, Brazil. The absorption spectrum from 260 nm to 650 nm indicated three main bands: the first around 270 nm related to phenolic acids and flavonoids; the second around 480 nm is associated with anthocyanins, and the third around 630 nm may be related to chlorophyll. The whole spectral range was analyzed using multivariate analysis methods and principal component analysis (PCA) was applied to determine the sample similarities and their relationship to geographic localization. Pre-processing methods, normalization, generalized least squares weighting -GLSW (classes based - color and region), mean center, and without GLSW, were compared. GLSW allowed propolis discrimination related to geographical origin and the corresponding photoacoustic signals to be verified. Propolis samples from southern and eastern Paran ' a showed photoacoustic signals of mean intensity, while samples from the western showed extreme minimum values. Interval partial least-squares-regression (iPLS) of the photoacoustic spectra and phenolic content had deter-mination coefficients equal to R2C = 0.959 (calibration) and R2P = 0.897 (prediction), root mean square errors RMSEC = 0.787% and RMSEP = 0.516% and residual prediction deviations RPDC = 5.10 and RPDP = 3.42, in the concentration range of 2.89 - 15.02 % (w/w). The results indicated that the spectrum obtained by photoacoustic spectroscopy can be successfully used to estimate the level of phenolics in raw propolis. This approach has the advantages of avoiding the step of propolis extraction and the need to carry out classical analytical protocols which are time-consuming and produce chemical residues.