The use of cellulose solvents is considered the most facile method to disrupt highly order hydrogen bonding of crystalline cellulose structure, increasing surface accessibility to enzymes, facilitating hydrolysis. Acknowledgement Enzymatic hydrolysis of corn stover (CS) Avicel + H 3 PO 4 Æ RAC COSLIF Avicel + ILÆ RAC IL Why cellulose solvents? Summary We compared two cellulose solvent-based lignocellulose pretreatments: COSLIF based on concentrated phosphoric acid and ionic liquid (IL) based on [BMIM]Cl. Enzymatic glucan digestibilities of COSLIF- and IL-pretreated corn stover were 96% and 52% by 5 filter paper units of cellulase per gram of glucan at hour 72, respectively. Regenerated amorphous (pure) cellulose from Avicel by COSLIF and IL had digestibilities of 100% and 92%, respectively. Our systematic experiments suggested that these large differences in enzymatic glucan digestibility of corn stover and Avicel pretreated by two cellulose solvents were attributed to the combinatory causes: (i) IL did not pretreat cellulose as efficiently as COSLIF based on enzymatic hydrolysis and cellulose accessibility to cellulase; (ii) the residual ionic liquid in the pretreated biomass inhibited cellulase activity; and (iii) the residual lignin on the cellulosic materials after the pretreatments decreased cellulase hydrolysis performance. Fig. 3. Hydrolysis of Avicel and RAC Fig. 2. Hydrolysis of corn stover Intact CS showed fibril structure (Fig. 2B), but COSLIF- and IL-pretreated CS did not show any fibril structure. CS COSLIF was hydrolyzed fast, and glucan digestibilities reached 93% at hour 72, while CS IL reached 55%. What caused such a large difference? (C) (B) before after before after 2 nd Hypothesis: IL inhibition (B) Intact (C) CS COSLIF (D) CS IL 0 12 24 36 48 60 72 0 20 40 60 80 100 Glucan digestibility (%) Time (h) CS COSLIF CS IL untreated (15 FPU) (A) 0 12 24 36 48 60 72 0 20 40 60 80 100 Glucan digestibility (%) Time (h) RAC COSLIF RAC IL Avicel (15 FPU) (A) 0 15 30 45 60 75 0 20 40 60 80 100 Glucan digestibility (%) Time (h) 0 g/L 1 gL 2 g/L 5 g/L 10 g/L (A) RAC COSLIF 0 12 24 36 48 60 72 0 20 40 60 80 100 0 g/L 1 g/L 2 g/L 5 g/L 10 g/L Glucan digestibility (%) Time (h) (B) Avicel 0 12 24 36 48 60 72 10 100 0 g/L 1 g/L 2 g/L 5 g/L 10 g/L (D) Avicel Normalized hydrolysis rate (%) Time (h) 0 12 24 36 48 60 72 10 100 0 g/L 1 g/L 2 g/L 5 g/L 10 g/L (C) RAC COSLIF Normalized hydrolysis rate (%) Time (h) Fig. 4. Influence of [BMIM]Cl on hydrolysis of RAC and Avicel 1 st Hypothesis: low pretreatment efficiency 3 rd Hypothesis: competitive lignin adsorption Material TSAC m 2 /g biomass CAC m 2 /g biomass NCAC m 2 /g biomass Pure substrate Avicel RAC COSLIF RAC IL 2.3 ± 0.1 19.1 ± 1.3 ND Lignocellulose Corn stover (CS) CS COSLIF CS IL CS DA 1.13 ± 0.01 14.4 ± 1.1 5.8 ±0.3 7.7 ± 0.6 0.42 ± 0.01 11.6 ± 0.8 5.0 ± 0.2 5.9 ± 0.3 0.71 ± 0.01 2.9 ± 0.2 0.73 ± 0.09 1.8 ± 0.6 Avicel can be completely dissolved in both cellulose solvents. Cellulose solutions appeared to be transparent (Figs. 3B & C). But RAC COSLIF yielded higher glucan digestibility than RAC IL . It was difficult to wash IL from RAC IL . ~0.1-1.0 g [BMIM]Cl remained on RAC IL . It was found that a small amount of [BMIM]Cl decreased hydrolysis rate and glucan conversion. Instantaneous hydrolysis rates of RAC and Avicel decreased drastically over time. 0 12 24 36 48 60 72 0 20 40 60 80 100 Glucan digestibility (%) Time (h) RAC IL RAC IL + lignin RAC IL + lignin IL Avicel Avicel + lignin (B) IL 0 12 24 36 48 60 72 0 20 40 60 80 100 Glucan digestibility (%) Time (h) RAC COSLIF RAC COSLIF + lignin RAC COSLIF + lignin COSLIF Avicel Avicel + lignin (A) COSLIF Fig. 5. Effect of isolated lignin on cellulose hydrolysis 0 12 24 36 48 60 72 0 20 40 60 80 100 Glucan digestibility (%) Time (h) RAC COSLIF (Avicel + lignin) COSLIF RAC COSLIF + lignin COSLIF (A) COSLIF 0 12 24 36 48 60 72 0 20 40 60 80 100 Glucan digestibility (%) Time (h) RAC IL (Avicel + lignin) IL RAC IL + lignin IL (B) IL 4 th Hypothesis: lignin redistribution Fig. 6. Effects of lignin redistribution (BMIMCl) Precipitation Tank ethanol W a s h e r - 1 Black Liquor W a s h e r - 2 Light Liquor Water lignin Corn stover Digester cellulose solvent CSTR amorphous cellulosic material ethanol water-soluble hemicellulose ethanol ethanol cellulose solvent Fig. 1. Cellulose solvent-based lignocellulose pretrement procedue H 3 PO 4 The negative effects of lignin addition were larger on RAC COSLIF than RAC IL . Lignin can redistribute on the surface of cellulose during pretreatment, which can block cellulose accessibility. We mixed lignin and Avicel prior to COSLIF and IL. there were no significant differences in hydrolysis rates and digestibilities between (Avicel+lignin) COSLIF and (Avicel+lignin) IL . Both H 3 PO 4 and [BMIM]Cl can dissolve cellulose and corn stover by disrupting ordered hydrogen bonds in cellulose fibers. A large difference in glucan digestibilities of CS COSLIF (96%) and CS IL (53%) was attributed to the combinatory effects: (1) IL did not pretreat cellulose as efficient as COSLIF, (2) residual IL inhibited cellulase activity, and (3) residual lignin on pretreated substrates decreased hydrolysis performance. But it was not due to lignin redistribution on the surface of cellulose. Cellulose accessibility to cellulase (CAC) was measured by a fusion protein, TGC. CS COSLIF had the highest CAC value of 11.6 m2/g biomass compared to those pretreated by IL and dilute acid (DA). At an enzyme loading of 5 FPU/g glucan, the enzymatic digestibilities increased as follows: COSLIF (96%) > DA (60%) > IL (53%) at hour 72, which is in good agreement with the CAC values in Table I. Discussion Table I. Surface accessibility assay Promising candidates Noppadon Sathitsuksanoh, Zhiguang Zhu, and Percival Zhang* Biological Systems Engineering Department, Virginia Tech, Blacksburg, VA, [email protected] Cellulose solvent-based lignocellulose pretreatment comparison: Concentrated phosphoric acid vs Ionic liquid