Enzyme and Microbial Technology 37 (2005) 195–204 A comparison between batch and fed-batch simultaneous saccharification and fermentation of steam pretreated spruce Andreas Rudolf * , Malek Alkasrawi, Guido Zacchi, Gunnar Lid´ en Department of Chemical Engineering, Lund University, Box 124, SE-221 00 Lund, Sweden Received 20 September 2004; received in revised form 10 January 2005; accepted 10 February 2005 Abstract In order to improve the process economy it is important to use as high dry matter content as possible in simultaneous saccharification and fermentation (SSF). However, too high dry matter content often gives rise to severe inhibition of the yeast metabolism, due to the increased levels of toxic compounds. The aim of the present work was to increase the fibrous content in SSF of steam pretreated spruce to 10% by adapting the yeast to the inhibitory substrate and by using a fed-batch process. Both batch and fed-batch approaches were evaluated. The fed-batch experiments were started with a batch fermentation containing 6% dry matter. Fibrous slurry from the pretreatment was then added four times during the first 24 h giving a final dry matter content corresponding to 10%. The yeast used in the fermentation was produced aerobically on the hemicellulose hydrolysate obtained from the pretreatment. SSF batch and fed-batch experiments with a cell mass concentration of 2, 3 and 5 g/L were carried out. When adapted yeast was used, the available hexoses were completely converted within 72 h and the final ethanol concentrations reached 40–44 g/L. No major differences in performance between batch and fed-batch were seen, but the ethanol productivity during the first 24 h was higher in the fed-batch SSF experiments, particularly during the experiments with a cell mass concentration of 2 and 3 g/L. © 2005 Elsevier Inc. All rights reserved. Keywords: Ethanol; Yeast; Fermentation; Enzymatic hydrolysis; Spruce 1. Introduction Reducing the production costs of ethanol produced from lignocellulose material is crucial in enabling its commercial- ization. One important factor for the production cost is the ethanol concentration in the liquid left after the saccharifica- tion and fermentation. This should be as high as possible [1] in order to minimize the energy costs of evaporation and dis- tillation. By recirculating some of the liquid from the fermen- tation back into the fermentor vessel, the final ethanol con- centration can be increased. However, compounds inhibitory to the yeast cells and enzymes, formed during the pretreat- ment of lignocellulose material [2–14], are thereby further concentrated, making a high degree of recirculation difficult [15,16]. Raising the dry matter content is another obvious way to reach a higher final ethanol concentration. However, * Corresponding author. Tel.: +46 46 222 82 50; fax: +46 46 14 91 56. E-mail address: andreas.Rudolf@chemeng.lth.se (A. Rudolf). the concentrations of compounds inhibitory to yeast and en- zymes are simultaneously increased. In addition, the rheolog- ical properties of a very dense fibrous suspension may cause mixing and heat transfer problems. Previously, when using compressed Baker’s yeast in Batch SSF with a dry matter content of 8% it proved difficult to achieve high ethanol yields (Alkasrawi et al., accepted for publication). The high concentrations of inhibitors in fermen- tations with a high dry matter content might be overcome by applying a fed-batch technique [17]. By adding the slurry to the fermentor vessel continuously or by scheduled additions, the inherent ability of the yeast to detoxify the substrate can be maintained. This approach has proved successful in fermen- tations of dilute-acid hydrolysates from softwood [18,19]. Furthermore, since added fibers are gradually degraded, the initial viscosity can be kept lower than in a batch process. Provided that it is possible to reduce the inhibition by ap- plying a fed-batch it could be possible to decrease the amount of yeast used in the SSF. Cultivation of yeast will consume 0141-0229/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.enzmictec.2005.02.013