In order to estimate an optimal fermentation time under ultrasonic exposure in this study, parameters such as ethanol concentration, ethanol volumetric productivity, ethanol yield and lactose consumption were investigated.

The maximum values of ethanol concentration and lactose consumption were achieved when the HRT was 36 h. Under the HRT of 36 h in the ultrasound-assisted fermentation, the average ethanol concentration of 26.30 g L-1, ethanol yield of 0.532 g g-1 lactose and lactose consumption of 98,9% were obtained (Fig. 9-11). Using S. cerevisiae without ultrasound exposure gave the results as 23,60 g L-1, 0.511 g g-1, 92,4%, respectively and the differences were statistically significant (p<0.05). Shortening the HRT to 24 h allowed remaining high ethanol yield of 0.520 g g-1 with sonicated S. cerevisiae, but in the control fermentation unit it was as low as 0.487 g g-1 (p<0.05). When the HRT was 12 h the ethanol yields were 0.318 and 0.365 g g-1 depending on using ultrasounds device (Fig. 11). From the economic viewpoint, shortening the fermentation time (HRT) could reduce costs of industrial ethanol production. The study showed that there is no need to extend the HRT over 36 h or more, because most of the lactose was converted into ethanol during 24 h (95.6% in the ultrasound-assisted fermentation. Nikolic et al. (2010) stated that optimal fermentation time for free and immobilized S. cerevisiae was 38 h. Ozmihci & Kargi (2008) studied ethanol production from cheese whey powder (CWP) solution containing 50 g sugar L-1 at six different HRTs varying between 17.6 and 64.4 h by Kluyveromyces marxianus strains. Percent sugar utilization,

12 24 36

HRT (h)

9. Effects of HRT and ultrasound irradiation on the ethanol concentration

HRT (h)

effluent ethanol concentration and ethanol yield increased with increasing HRT from 17.6 to 50 h. Further increasing in HRT to 64.4 h resulted in decrease of the analyzed parameters. Moreover, the time for fermentation decreased at higher initial substrate concentration (Guimaraes et al., 2008a; Nikolic et al., 2010; Ozmihci & Kargi, 2008). According to Guimaraes et al. (2008a) the fermentations with 50-150 g lactose L-1 reached completion in about the same time of 27 h but the maximum ethanol concentration increased linearly with increasing initial lactose concentration from 6.5 g ethanol L-1 with 20 g lactose L-1 to 57 g L-1 with 200 g L-1. They also stated that increasing lactose concentration led to incomplete fermentation and impair the fermentation due to nutrient limitation.

Interestingly, the volumetric productivities of ethanol decreased at longer HRT (Table 3). Maximum productivity of ethanol of 1.060 g L-1 h-1 was observed under the HRT of 12 h when the culture has been sonicated and 0.908 g L-1 h-1 under the HRT of 24 h in the fermentation process without ultrasound irradiation (p<0.05). The volumetric ethanol productivity in the ultrasound-assisted fermentation obtained in this work was higher than that reported for batch or fed-batch fermentations with S. cerevisiae strains: 0.3 g L-1 h-1 (Rubio-Texeira et al., 1998), 0.46 g L-1 h-1 (Guimaraes et al., 2008b), 0.14 — 0.6 g L-1 h-1 (Ramakrishnan & Hartley, 1993), 1 g L-1 h-1 (Compagno et al., 1995). Ozmihci & Kargi (2007) using Kluyveromyces marxianus to ferment concentrated cheese whey powder solution obtained higher volumetric ethanol productivity over 2 g L-1 h-1, but after 120 h fermentation.