No abstract

No abstract

No abstract

The authenticity and composition of commercial essential oils requires strict quality control. Due to the importance of α-pinene containing essential oils, a rapid and efficient method for quantification of this terpene in oils of eucalyptus, pink pepper and turpentine using (1)H NMR was developed and validated. All evaluated parameters (selectivity, linearity, accuracy/precision, repeatability, robustness, stability of analyte and internal standard in solutions) showed satisfactory results. The limit of detection (LOD) and limit of quantification (LOQ) were 0.1 and 2.5mg respectively. These values indicated that α-pinene was detected in 35 mg samples containing at least 0.3% of this compound. In addition, a minimum of 8% of α-pinene in the sample was required for quantification. Furthermore, the standard deviations found in the (1)H NMR methodology were less than 1% and were lower than those obtained by gas chromatographic analysis. Statistical tests have shown that the results obtained by (1)H NMR methodology are similar to those obtained by GC-FID technique using external and internal standardization and normalization within 95% confidence. R&R values lower than 10% have shown that all the methods are appropriate and the (1)H NMR method is suitable for quantification of α-pinene in samples of essential oils since this method possessed the smallest R&R (1.81) value.

This paper describes how artificial neural networks can be used to classify multivariate data. Two types of neural networks were applied: a counter propagation neural network (CP-ANN) and a radial basis function network (RBFN). These strategies were used to classify soil samples from different geographical regions in Brazil by means of their near-infrared (diffuse reflectance) spectra. The results were better with CP-ANN (classification error 8.6%) than with RBFN (classification error 11.0%).