The development of cellulosic ethanol based biorefineries has been largely impeded due to the native recalcitrance of plant cell walls to chemical pretreatments and enzymatic hydrolysis. Lignin-carbohydrate complex (LCC) ester linkages and cellulose crystallinity are major impediments to efficient cell wall deconstruction. Ammonia based pretreatments like AFEX (or Ammonia Fiber Expansion) are thought to overcome cell wall recalcitrance through ammonolysis of ester linkages and de-crystallization of cellulose. However, the major ammonolysis end-products and pretreatment induced cell wall based ultrastructural changes have never before been determined for AFEX treated lignocellulosics. Previous research has focused mostly on acidic pretreatments that extensively extract hemicellulose and lignin, hence minimizing the requirement for optimal cellulase and hemicellulase based enzyme cocktails. The goal of this study was to elucidate the mechanism of ammonia based pretreatments on lignocellulosic plant cell walls (e.g. corn stover, cellulose) and examine how that influences enzymatic activity.

In this study, it was found that ammonolysis of corn stover LCC ester linkages resulted in the extraction and re-deposition of decomposition products (e.g. amides, arabinoxylan oligomers, lignin phenolics) on outer cell wall surfaces. These amorphously shaped deposits were characterized by confocal fluorescence, atomic force and scanning electron based microscopic techniques. Cleavage of LCC linkages during AFEX resulted in the formation of a nano-porous network that likely enhanced enzyme accessibility. The shape, size (e.g. pore width ranged between 10-1000 nm) and spatial distribution of the porous network was dependent on cell wall composition (e.g. middle lamella vs. secondary walls) and ammonia pretreatment conditions. Pretreated cell wall cumulative porosity (i.e. pore surface area per unit cell wall volume), estimated via analysis of transmission electron microscopy based tomograms (0.005-0.05 nm ²/nm³), was linearly correlated to the total LCC ester linkages cleaved. Closer examination of cellulose crystal structure revealed formation of cellulose III allomorph upon treatment with anhydrous liquid ammonia (but not with ammonium hydroxide). Purified endoglucanase I (Glycosyl hydrolase or GH family 7), isolated from T. reesei, was found to have two-fold higher specific activity on cellulose III versus untreated cellulose. Cellulose within ammonia treated cell walls (ACW) was more readily accessible to T. reesei based cellobiohydrolases (GH 6, 7) compared to untreated cell walls. Further addition of endoxylanases (GH 11), β-xylosidase (GH 3) and α-arabinofuranosidase (GH 51) increased both glucan and xylan digestibility for ACW. Fourteen commercially available crude enzymes (ranging in diverse GH activities based on proteomics analysis) were optimally blended, to maximize ACW digestibility, using a high-throughput microassay. The optimal blend gave 90% total sugar conversion at 2-fold lower enzyme loading (i.e. 16.5 mg/g glucan) compared to commercial cocktails (e.g. Spezyme CP + βG rich cocktail gave 50% total sugar conversion). In summary, formation of cellulose III allomorph coupled with rationally tailored enzyme mixtures for ammonia treated lignocellulosic biomass could significantly reduce cellulosic ethanol production costs.