Journal
JOURNAL OF THE INSTITUTE OF BREWING
Volume 112, Issue 1, Pages 36-40Publisher
INST BREWING
DOI: 10.1002/j.2050-0416.2006.tb00705.x
Keywords
artificial neural network; chemometrics; multivariate modelling; quality control; staling markers
Categories
Ask authors/readers for more resources
Key compounds in lager staling include furfural, hexanal, 5-hydroxymethyl furfural (5-HMF), and trans-2-nonenal. Quantitative data of headspace concentration in two lagers - one premium at 5% (abv), the other a standard product at 4% (abv) were obtained by solid phase microextraction (SPME) followed by gas chromatography using a mass selective detector (GCMS). The concentrations of the aldehydes were used to predict overall stale scoring from sensory assessor data, of lagers stored at 4, 12, 30, and 37 degrees C for 7, 14, 21 and 28 days. Concentrations of all four aldehydes increased with time of storage and with higher temperatures. Correlation coefficients for prediction of staleness in the premium lager were similar at 0.81 and 0.84 for partial least square regression (PLS1) and artificial neural network (ANN) modelling respectively, and the latter showed a lower root mean square error (RMS error). For the standard product, the correlation coefficients were 0.72 and 0.86, with ANN showing lower RMS error respectively. In both PLS models, E-2-nonenal had high regression coefficients and 5-HMF lower coefficients. Furfural and hexanal differed in contributions to the lagers.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available