Abstract Mackenzie Aragon | Robert Coleman | Roger Boulton | Thomas S. Collins

Influence of Grain Type and Surface Area on Extraction of Alternative Oak in a Model Wine Solution

Mackenzie Aragon,* Robert Coleman, Roger Boulton, and Thomas S. Collins
*Washington State University, 2455 George Washington Way, APT F230, 
Richland, WA, 99354 (mackenzie.aragon@wsu.edu)

One advantage of using alternative oak products is more rapid extraction than from traditional oak barrels, which is typically attributed to increased surface/volume ratio, without regard to grain type (cross- versus end-grain). This work evaluated how surface area (SA) and grain type influence extraction of oak aroma compounds. Medium plus convection-toasted staves were cut sequentially in half by length (25.4, 12.7, 6.35, 3.18 and 1.59 cm), resulting in treatments with the same total weight and volume of oak, but increased SA and end-grain to cross-grain ratio with shorter segments. All treatments used a dosage rate of 20 g/L and were extracted in model wine (15% v/v ethanol) for 240 days. Samples were analyzed using headspace solid-phase microextraction gas chromatography-mass spectrometry (HS SPME GC-MS) to determine concentrations of furfurals, guaiacols, and lactones.

If the increased extraction rate in oak alternatives is due only to increased SA, then the extract concentration should increase proportionally between treatments with the SA increase. Instead, the extraction rate and the concentration increased as both the SA and the end-grain to cross-grain ratio increased. When the shortest segments are compared to the longest segments (45% more SA) at day 10 of the extraction, concentration increases of 130% to 15,000% (varying by compound) were observed. This suggests that increased exposure to end-grain is more important to rapid extraction than SA, as the additional end-grain accounts for at least two-thirds of the additional extraction from shorter segments. A first-order kinetic model for each compound at each SA and grain ratio was fitted to the raw data using the differential evolution parameter estimation routine (Peterson and Ulrich 2011). The outcome of this model supports the importance of grain ratio on both extraction rate and concentration.

Funding Support: Scott Labs