Comprehensive two-dimensional gas chromatography for fingerprint pattern recognition in cachac?a production

Zenilda Cardeal, P P de Souza, Marco Gomes da Silva, Philip Marriott

Research output: Contribution to journalArticleResearchpeer-review

71 Citations (Scopus)

Abstract

Cachaca samples were studied by means of comprehensive two-dimensional gas chromatography and time-of-flight mass spectrometry (GC x GC/TOFMS) during the fermentation process and after ageing in different wood materials. The analyses of the aroma compounds were performed after headspace-solid phase microextraction method (HS-SPME) using an 85 mu m polyacrylate (PA) fibre. Fingerprint monitoring of the distillation process allowed the easy determination of the turning points of the process and high-resolution comparison of cabeca (head), coracao (core) and cauda (tail) fractions. The ageing process in different wood materials was well characterised through fingerprint similarity observations; in the absence of a suitable metric for expressing the overall similarity, here we use a visual and retention time comparison to identify co-incident peaks and those that differ between samples. For quality control purposes, a simple observation of the contour plots obtained can thus allow the identification of the type of wood used in the ageing process, and the process of ageing, without further statistical treatment or peak identifications. In this manner, peaks, which discriminated most between the different mixtures studied were readily found, i.e. unique compounds were identified in each stage of the distillation process. Approximate first dimension linear retention indices (LRI) for these identified compounds were calculated in a bi-dimensional polar/non-polar column set in the GC x GC experiment and were used in conjunction with mass spectral library searching for tentative identification. Along the progression of the distillation process, 70 compounds appear to visually discriminate between samples and their retention indices are indicated, presenting good correlation with literature data.
Original languageEnglish
Pages (from-to)793 - 799
Number of pages7
JournalTalanta
Volume74
Issue number4
DOIs
Publication statusPublished - 2008
Externally publishedYes

Cite this