No abstract

The potential of NIR reflectance spectroscopy for discriminating between pure Arabica and pure Robusta coffees and blends of these two was investigated. Studies were performed on whole and ground beans using a factorial discriminant procedure. For whole beans, in the absence of blended samples, a correct classification rate of 96.2% was achieved. Inclusion of blended samples reduced this figure to between 82.9 and 87.6%. In the case of ground samples, including blends, a correct identification rate of 83.02% was achieved. The molecular basis for discrimination is discussed.

No abstract

The aromatic compositions of a marine alga (Fucus serratus) and its maceration were determined. Sixteen major volatile compounds were selected as being representative of algal aroma components. Pervaporation performance was investigated using a model multicomponent solution and low aroma compound¯uxes were obtained. Interesting selectivities were calculated showing the concentration potential of pervaporation. Comparison with maceration pervaporation was made and showed similar selectivity values. Finally, the sensorial quality of the maceration pervaporate was evaluated regarding the operating parameters of pervaporation. The permeate pressure had a statistically signi®cant effect upon aromatic quality.

No abstract