2%); this is further shown in Table 3. However, it should be further noted that in both cases the performance of the APCI-MS as a tool for geographical provenance determination was very good considering the high intrinsic variability due to the use of commercial samples. LY294002 Whilst the use of commercial samples does allow the inclusion of true sample variability, it does not permit strict control of process parameters that support a mechanistic explanation of the model (e.g. cultivation and irrigation practices, environmental factors, edaphological parameters, post-harvesting practices). In both internal and external validation datasets the samples originating from New Zealand
were all successfully classified, and of the total 135 samples only 4 were misclassified, resulting in an error rate of <3%. There was a similar correlation of m/z to principle components, to that observed previously ( Fig. 4b). More specifically, the first axis is proposed to be related to alkyl-esters (m/z 61, 75, 85, 89,
103, 117, 131, 145) and dehydrated alcohols (i.e. m/z 85 for 1-hexanol, m/z 57 for 1-butanol) in the form of fragments or parent ions. The second most powerful discriminating factor is shown on PC 2 and was found to associated with the green-grassy odour like volatiles such as 1-hexanal and trans-2-hexenal (m/z 101 and 99, respectively), or 1-hexanal and cis-hex-3-en-ol selleck chemical (m/z 83). Thus, complete discrimination between New Zealand and South Africa juices appear to be dependent on the ester-related flavour notes (fruity–flowery), whilst the Chilean samples appear to be discriminated by moderate ester concentration and low amounts of green-grassy flavour type volatiles. Finally, it should be noted that the variability of the New Zealand and South Africa labelled juices based on the green-grassy flavour criterion was quite high which would indicate differences in the ripening level of the sampled apples. In conclusion, a PLS-DA chemometric approach was demonstrated to be a viable tool for the interpretation of raw APCI-MS data. The models generated were robust enough to reliably discriminate (100% correct
classification with external validation set) Tolmetin apple juices prepared from Braeburn, Golden Delicious, Granny Smith, Jazz and Pink Lady varieties, furthermore developments on the model allowed the reliable (94.2% correct classification with external validation set) discrimination of the geographical provenance of monovarietal clarified apples from Chile, New Zealand and South Africa. “
“Coffee is one of the most valuable basic products, constituting the second major commodity just after oil (ICO, 2012 and Nabais et al., 2008). According to the International Coffee Organization – ICO (2012), the total coffee production in crop year 2011/2012 was about 131.3 million bags (each bag weighing 60 kg), with approximately 33.1% produced in Brazil.