Issue link: http://winesandvines.uberflip.com/i/235959
w i n e M A K IN G Compound Furfural 5-methylfurfural Vanillin Syringaldehyde Vanillic Acid Gallic Acid Longitudinal Exposure Extraction Rate (r 2 value) 312 (0.91) 58 (0.97) 28 (0.97) 89 (0.97) 91 (0.97) 73 (0.95) Radial Exposure Extraction Rate (r 2 value) 189 (0.99) 25 (0.99) 12 (0.79) 34 (0.76) 34 (0.95) 18 (0.93) Table I: Initial extraction rates for compounds measured by HPLC during extraction based on slope of concentration to time graph over the first 21 days of extraction. Units are μg/L-day. Furfural 5-methylfurfural Vanillin Syringaldehyde Vanillic Acid Gallic Acid Guaiacol 4-methylguaiacol 4-ethylguaiacol Eugenol Longitudinal Radial p-value Compound tion very similar to theExposure Final Exposure Finalis both a smaller molecule and 10-week sample Furfural (1,257±38 μg/L at 10 Concentration weeks, 1,286±44 more water-soluble than vanillin, which Concentration μg/L at 12 weeks). could explain some of the difference in Furfural 13.0±0.3 ppm 12.0±0.9 ppm 0.9 The results demonstrate that furfural is extraction 5-methylfurfural 2.10±0.07 ppm 1.61±0.1 ppm rates between these two com0.5 completely extracted from cross-cut pounds rather than any differences in Vanillin 1.22±0.10 ppm 0.69±0.15 ppm 0.05* exposed oak in approximately two distribution through the wood. Syringaldehyde 4.7±0.3 ppm 2.2±0.4 ppm 0.0008* weeks, while more than 10 weeks is It is important to note that furfural Vanillic Acid 4.0±0.5 ppm 1.8±0.4 ppm 0.002* required for extraction of furfural from extraction from cross-cut exposed wood Gallic Acid 3.8±0.6 ppmis sig1.4±0.3 ppm 0.001* parallel-exposed oak. This result was complete after about two weeks, Guaiacol 49±9 ppb 33±1while vanillin continued increasing ppb 0.04* nificant considering that the relative sur4-methylguaiacol 17±2 ppb oak is ppb 8±2 through the 0.01* week and was also face area of the parallel-exposed fifth 4-ethylguaiacol 85±7 ppb 51±6extracted into the second solution. This ppb 0.003* greater than the cross-cut exposed oak Eugenol 10.2±0.7 ppb 5.8±0.4 ppb 0.0008* (52 cm2 compared to 40 cm2). We thereeffect indicates that re-use of oak alternafore conclude that the extraction of furfutives would result in different flavor proral from the cross-cut axis of oak occurs files due to the differences in extraction at a rate at least five times greater than rates for these two compounds. extraction from the parallel axis. For example, a five-week exposure of the cross-cut exposed oak to a wine soluVanillin (Experiment 1) tion would result in about 1,200 μg/L of Figure 3 is a graph of the concentration of furfural and almost 400 μg/L of vanillin vanillin in the extraction solution in a in wine after the first use. Upon re-use of 10-week experiment. Only the cross-cut the oak, a five-week exposure to fresh exposed sample is plotted as the parallelwine would result in only minimal furfuexposed flask had no measurable vanillin ral (less than 300 μg/L) and about 400 (less than 100 μg/L) during the entire μg/L of vanillin. experiment. The results also demonstrate that vanilThe cross-cut exposed sample showed an lin is extracted much more efficiently in increasing vanillin concentration through cross-cut exposed wood than parallelthe first five weeks, when the extraction exposed wood. This result demonstrates solution was changed (due to the complete the benefit of cross-cut grain exposure in extraction of furfural as mentioned above). barrel alternatives. In our study, the The vanillin concentration did rebound to cross-cut exposure provides a much more approximately the same level in the secrapid extraction of the measured flavor ond extraction solution as it reached in the compounds into the wine compared to first. The slope of the vanillin graph is parallel exposure. similar in both five-week extractions. The Experiment 2 vanillin results demonstrate that this comAfter one week the model wine solutions pound is extracted much more readily containing the cross-cut exposed oak had from cross-cut exposed oak than from a noticeably darker appearance than the parallel-exposed oak. solutions containing the parallel-exposed Comparison of furfural and vanillin oak indicating a higher level of extracted (Experiment 1) phenolics. Both the visible absorbance at It is clear that furfural extracted more rap465 nm and the six HPLC-measured anaidly than vanillin from both cross-cut lyte compound concentrations were exposed and parallel-exposed wood. higher in the model wine solution con24 p racti c al w i ne ry & v i n e yard JANUARY 20 14 Radial Exposure Final Concentration 12.0±0.9 ppm 1.61±0.1 ppm 0.69±0.15 ppm 2.2±0.4 ppm 1.8±0.4 ppm 1.4±0.3 ppm 33±1 ppb 8±2 ppb 51±6 ppb 5.8±0.4 ppb p-value 0.9 0.5 0.05* 0.0008* 0.002* 0.001* 0.04* 0.01* 0.003* 0.0008* Table II: Summary of final concentrations of flavor molecules in model wine solution. Each number represents the average of three model wine solutions. Significant differences are indicated with a * (p-value 0.05 or less). taining longitudinal-exposed oak after only one day of extraction. Figure 4 is a plot of the absorbance of the model wine solution at 465 nm versus time of extraction. The average absorbance for each of the three replicates is plotted with its standard deviation. Both solutions appear to reach a steady-state absorbance well before the end of the 14-week extraction period. The absorbance of the model wine containing cross-cut exposed oak was more than three times that of the solution containing parallel-exposed oak. 3000 2500 Concentration (ppb) Compound Longitudinal Exposure Final Concentration 13.0±0.3 ppm 2.10±0.07 ppm 1.22±0.10 ppm 4.7±0.3 ppm 4.0±0.5 ppm 3.8±0.6 ppm 49±9 ppb 17±2 ppb 85±7 ppb 10.2±0.7 ppb 2000 1500 1000 500 0 0 5 10 15 20 25 -500 Time (days) Figure 5: Concentration of vanillic acid to time for the first 21 days of extraction (Experiment 2). ∆ = concentrations in solution exposed to parallel-grain oak, X = cross-cutexposed oak solutions. Slope of cross-cut data = 91μg/L-day, slope of parallel data = 34μg/L-day. Rate of extraction (Experiment 2) In order to determine the rate of extraction of the analyte compounds, the slope of the concentration versus time graph was measured using the first five data points (day one, three, seven, 14 and 21) for each analyte determined by HPLC. Figure 5 is one example of the data used for this analysis, showing the concentration of vanillic acid related to time for each type of oak grain exposure.