Quantification of oak wood extractives via gas chromatography - mass spectrometry and subsequent calibration of near infrared reflectance to predict the Canadian whisky aging process
MetadataShow full item record
The determination of oak wood contribution to the Canadian whisky ageing process was studied. A direct injection – gas chromatography mass spectrometry method was developed and optimized to accurately monitor thermally degraded wood extracts derived from hemicellulose, cellulose, and lignin. Thermally degraded wood extracts included furfural, 5-methylfurfural, furfuryl alcohol, cyclotene, guaiacol, cis-whisky lactone, creosol, trans-whisky lactone, maltol, 4-ethylguaiacol, eugenol, 4-ethylphenol, syringol, 5-hydroxymethylfurfural, vanillin, and syringaldehyde. A three year ageing process of base Canadian whisky in brand new casks with 2 mm and 4 mm depth char, once used American Bourbon casks, and refurbished rechar casks were monitored over time for the thermally degraded wood extracts and compared. New casks had 80000 ppb of extract/LAA (litres of absolute alcohol), once-used AB casks had 20000 ppb of extract/LAA, and refurbished – rechar casks had 18000 ppb of extract/LAA at the end of the three year period. Extractions of wood compounds were logarithmic profiles. The time taken to extract half the total wood components of the three year ageing process was 30 - 50 days. To predict the ageing process in Canadian whisky the chemical structure of oak wood staves were determined using various alkaline and acid gravimetric digestions, and the results were calibrated with a fibre optic accessory from an NIR. The calibration models developed were neutral detergent solubles (RMSECV = 2.65), hemicellulose (RMSECV = 2.13), lignin (RMSECV = 2.42), cellulose (RMSECV = 7.22), and char (RMSECV = 7.07). The models developed were adequate for screening purposes with RPD values of 1.53 – 3.74. A 60 day disc ageing study for bench top ageing experiments was developed to correlate the compound changes over time and the wood derived extracts in the surrounding spirit. It was determined that hemicellulose, cellulose, and lignin derived compounds extract at different rates, and the error of the developed calibration models was too great in order to correlate loss in hemicellulose, lignin, and cellulose to the amount of wood extract that entered the whisky. The 60 day disc protocol was used to directly calibrate total wood extract in the spirit to the NIR probe. It was determined that NIR could be successfully used to predict the thermally degraded wood extract available on the surface layer below the char on a bench top level. The calibration model had a RMSECV of 914 ppb, R2 of 0.94, RPD of 3.98, and a RER of 32.53.