Effect of hemicellulose and lignin removal on enzymatic hydrolysis of steam pretreated corn stover

Abstract

Ethanol can be produced from lignocellulosic biomass using steam pretreatment followed by enzymatic hydrolysis and fermentation. The sugar yields, from both hemicellulose and cellulose are critical parameters for an economically-feasible ethanol production process. This study shows that a near-theoretical glucose yield (96–104%) from acid-catalysed steam pretreated corn stover can be obtained if xylanases are used to supplement cellulases during hydrolysis. Xylanases hydrolyse residual hemicellulose, thereby improving the access of enzymes to cellulose. Under these conditions, xylose yields reached 70–74%. When pre-treatment severity was reduced by using autocatalysis instead of acid-catalysed steam pretreatment, xylose yields were increased to 80–86%. Partial delignification of pretreated material was also evaluated as a way to increase the overall sugar yield. The overall glucose yield increased slightly due to delignification but the overall xylose yield decreased due to hemicellulose loss in the delignification step.

The data also demonstrate that steam pretreatment is a robust process: corn stover from Europe and North America showed only minor differences in behaviour.

Introduction

Bioethanol can be produced from lignocellulosic material using pretreatment followed by enzymatic hydrolysis and fermentation (Fan et al., 1982, Saddler et al., 1993, Sun and Cheng, 2002, Mosier et al., 2005). In order to obtain a high overall ethanol yield, the pretreatment step should improve the accessibility of the cellulose component to hydrolytic enzymes while avoiding degradation of solubilised hemicellulose and cellulose. Efficient sugar-to-ethanol conversion is necessary in order to achieve an economically feasible production process because the raw material represents a large part of the total costs (McMillan, 1996). Sugar degradation not only decreases the final ethanol yield but also results in degradation products that are inhibitory to the yeast used in the subsequent fermentation (Palmqvist et al., 1996).

Several different pretreatment methods can be used to facilitate the enzymatic hydrolysis of lignocellulosic material (Sun and Cheng, 2002, Mosier et al., 2005). One of the most thoroughly investigated methods is steam pretreatment using an acid catalyst (Galbe and Zacchi, 2002). Steam pretreatment of corn stover at 190 °C for 5 min using SO₂ as acid catalyst has been shown to give high sugar yields (almost 90% overall glucose yield and almost 80% overall xylose yield) after 72 h enzymatic hydrolysis (Öhgren et al., 2005).

The steam pretreated slurry was shown to be fermentable using ordinary baker’s yeast (Saccharomyces cerevisiae), using either simultaneous saccharification and fermentation (SSF) or separate hydrolysis and fermentation (SHF) (Öhgren et al., 2006a, Bura, 2005) but since corn stover contains large amounts of xylose, a pentose sugar that is not fermentable using ordinary baker’s yeast, a recombinant xylose fermenting yeast would be preferable to use in order to improve ethanol yield. However, recombinant S. cerevisiae strains are more sensitive to metabolic inhibitors generated during steam pretreatment (Öhgren et al., 2006b). Unfortunately, attempts to mitigate inhibitor formation by reducing pretreatment severity usually result in unacceptably low enzymatic digestibility.

In order to obtain fast enzymatic hydrolysis of biomass with a high sugar yield (for both hexoses and pentoses), the two main protective coats around cellulose – hemicellulose and lignin – need to be removed or altered without degrading the hemicellulose sugars. Hemicellulose forms a physical barrier around the cellulose (Hsu, 1996). At high severity pretreatment the hemicellulose is hydrolysed to monomeric sugars and then further degraded. In low severity pretreatments however, the recovery of hemicellulose sugars is high but residual hemicellulose hinders enzyme access to cellulose. Lignin also forms a protective physical barrier to enzymatic attack and is not extensively removed by steam pretreatment. However, pretreatment results in a partial melting and coagulation that reduces steric interference (Hsu, 1996). After steam pretreatment, inhibition of hydrolysis is largely due to non-specific hydrophobic binding of lignin to the cellulose binding domain of the enzymes (Palonen et al., 2004).

In an attempt to overcome the obstacles to complete enzymatic hydrolysis, two methods were evaluated in this study. Additional xylanase was added to the cellulase mixture in order to increase the hemicellulose hydrolysis and thus increase cellulase accessibility. Increased hydrolysis rate should also lead to decreased hydrolysis time and hence reduced process cost (Wingren et al., 2003). A partial delignification step, using ethanol organosolv pulping, was also performed after pretreatment in order to decrease the lignin content and thus reduce non-productive enzyme–lignin-binding. A previous study demonstrated improved sugar yield and hydrolysis rate following delignification of steam pretreated material (Pan et al., 2004). These strategies were also applied to materials pretreated under milder conditions in an attempt to improve xylose recovery while maintaining high hydrolysis yield for glucose.

The severity of steam pretreatment can be decreased by reducing the temperature (pressure) or by decreasing the amount of acid catalyst added. Temperature (pressure) reduction saves energy and material costs while adding less acid catalyst decreases the production of pollutants and lowers corrosion problems (Garrote et al., 1999). Steam pretreatment is possible without added catalyst as a result of autocatalysis: hydronium ions from water and compounds generated in situ (e.g., acetic, uronic and phenolic acids) catalyse the hemicellulose hydrolysis (Garrote et al., 1999). The efficacy of autohydrolysis will thus, to some extent, depend on the amount of acids generated in situ.

Corn stover, an abundant agricultural by-product (Kim and Dale, 2004) in both Europe and North America, was used as raw material in this study. Previous studies have reported a large diversity in corn stover composition depending on both genetics and environmental factors (Thomas, 2003, personal communications). In order to investigate the robustness of the steam pre-treatment process, corn stover from Italy and North America was compared in this study.

The overall objective was to obtain high yields of both hexose and pentose sugars. Three different pretreatment conditions followed by enzymatic hydrolysis were investigated. The optimal pretreatment condition chosen was based on previous data (Öhgren et al., 2005) Decreased temperature and autocatalysis were used to reduce pre-treatment severity. Hydrolysis was performed using either standard conditions (15 FPU cellulase/g cellulose) or with xylanse supplementation (0.06 g protein/g cellulose). Control reactions containing BSA or additional cellulase at the same protein loading were also included. In further experiments, delignification was performed prior to the enzymatic hydrolysis under standard conditions.