The effect of the barrier material seemed to be overridden by the action of OS at the beginning of the storage, but once the scavenging capacity started to deplete, the sample packaged with a lower O2 barrier started to show a slow increase in microbial growth. Nonetheless, the growth rate towards the end of the storage testwas slower than that of the control sample. Thiswas caused by the effect of high headspace CO2. Although the application of EE can suppress the initial growth, the choice of barrier material was also relevant to the overall effectiveness of extending the shelf life of the cake. The results of microbial testing correlated to the decrease of ethanol concentration (Fig. 3). The efficiency of nylonpackaged samples was depleted by the dissipation of ethanol through the nylon bag. PVDC-coated nylon had a much lower ethanol transmission rate (less than 1 g/m2.day at 40 C) than that of nylon (15e30 g/m2.day at 40 C) (Yam, 2009).