Conversion of Lignocellulosic Biomass into Monomeric Sugars and Levulinic Acid: Alkaline Pretreatment, Enzymatic Hydrolysis, and Decrystallization

Abstract

The technical feasibility of applying uniform pretreatment conditions across the different switchgrass varieties to achieve similar sugar yields was investigated. Four different types of switchgrass species including, Shawnee, Alamo-I and Alamo-II, and Dacotah were investigated using aqueous ammonia as a pretreatment reagent. A majority of the pretreatment experiments were carried out using Soaking in Aqueous Ammonia (SAA) method; while a selected few pretreatment experiments with Dacotah were carried out using Ammonia Recycle Percolation (ARP) method. SAA is a batch process, whereas ARP is a flow-through semi-batch process. The pretreated solids were subjected to enzymatic hydrolysis using 15 FPU/g-glucan of Spezyme-CP and 30 CBU/g-glucan of Novozymes-188. Hydrolysis results show that with the similar pretreatment conditions (90°C, 15% NH4OH, liquid to solid ratio of 9, and 24 hr), Alamo-I, Alamo-II, and Shawnee resulted in substantially higher glucan/xylan (as high as 19%/13%) digestibility than the Dacotah. The optimum pretreatment conditions for Shawnee, Alamo-I, and Alamo-II were: 90°C, 15% NH4OH, liquid to solid ratio of 9, and 24 hr. Increase in SAA pretreatment severity and applying ARP pretreatment increased the glucan/xylan digestibility of Dacotah, but also caused significant loss of carbohydrate content during pretreatment. The majority of results available in the literature for pretreated solids report high saccharification yields using uneconomically high enzyme loadings. Therefore, the goal of this work was to create a defined enzyme mixture that results in high yields of six and five carbon sugars from the pretreated switchgrass with the least amount of enzyme loading. The enzymatic hydrolysis experiments were carried out by applying various mixtures and levels of commercial enzymes including, Spezyme-CP (cellulase), Novozyme-188 (β-glucosidase), Multifect Xylanase (xylanase), and Multifect Pectinase (pectinase) with the solids produced from SAA pretreated Dacotah. Addition of β-glucosidase increased the glucan and xylan digestibilities with all levels of enzyme loading throughout the hydrolysis. Glucan and xylan digestibilities were rapidly increased approximately up to 25 mg protein/g-glucan with and without addition of β-glucosidase, after which the increase was gradual. The increase in the total protein loading of cellulase and β-glucosidase from 25 to 32 mg protein/g-glucan only increased the glucan/xylan digestibility by 4%/3%, whereas by supplementing same amount of xylanase and pectinase increased the glucan/xylan digestibility by 21%/22% and 10%/19%, respectively. Addition of β-glucosidase decreased the cellobiose, glucose and xylose oligomers. Supplementation of xylanase/pectinase only reduced xylose oligomer at 72 hr hydrolysis period. Use of surface active additives is expected to reduce the enzyme dosage in the bioconversion processes by 20-40%, a significant economic benefit. Therefore, in this work to evaluate the effect of additives on enzymatic hydrolysis, different water soluble polymers were studied including, non-ionic (polyvinyl alcohol, polyvinyl pyrrolidone, and polyethylene glycol) and ionic (cationic polyacrylamide and polyethylene imine) polymers. In addition, to evaluate the effect of presence of lignin, various feedstocks including, Solka Floc and hardwood pulp (lignin free substrates), dilute sulfuric acid pretreated corn stover, dilute sulfuric acid pretreated switchgrass, aqueous ammonia pretreated switchgrass (lignin containing substrates) were used. The enzymatic hydrolysis experiments were carried out by applying 5 FPU/ g-glucan of C Tec 2. Hydrolysis results indicate that with the supplementation of 10 mg polymer/g-substrate, glucan/xylan digestibilities were increased by 10-30% depending on type of the substrate and polymer. The non-ionic polymer (PVP, PVA and PEG) addition resulted in significantly higher hydrolysis yields than the ionic polymers (PEI and C-PAM) for all substrates. The main reasons for improvement in digestibility with polymer addition are improvement in cellulase activity and reduction in unproductive binding of cellulase to lignin. Sulfur containing reagents are highly reactive with lignin, thus can serve as delignification enhancers. Therefore, the effect of adding small amount of sodium sulfide in alkaline pretreatment of Dacotah switchgrass and palm residue was investigated. The alkaline reagents tested were either recoverable ones (ammonia, sodium carbonate) or inexpensive one (lime). Pretreatments were primarily carried out in batch mode under low severity conditions: 90-120°C, 12-24 hr treatment time, 0-15% alkali, 0-3% sodium sulfide, and liquid to solid ratio of 9. The enzymatic hydrolysis experiments were carried out by applying 5 FPU/ g-glucan of C Tec 2. Addition of sodium sulfide did not affect the carbohydrate content, but increased the delignification by 14-42%. Addition of sodium sulfide substantially increased the digestibility for all the alkali treatments: 15-45% in glucan digestibility and 10-40% in xylan digestibility depending on type of substrate and pretreatment severity. Levulinic acid (LA) is an ideal platform chemical that can be utilized to produce a number of bio-chemicals. However, the majority of the results reported in the literature are about LA production from monomeric sugars/lignocellulosics using dilute sulfuric concentrations (1-5% (w/v)), low substrate loadings (1-5% (w/v)), and at high temperatures (160-220oC). Thus, this work mainly focused on evaluating the technical feasibility of production of LA using high substrate loadings (10-20%), moderate acid concentrations (10-25%), and at low temperatures (80-140oC). LA production from different soluble sugars (glucose, fructose, sucrose, galactose, and mannose), lignocellulosic substrates (Solka Floc, corn stover, pretreated corn stover, pine, and municipal solid waste), and starch was studied. Study shows that it is technically possible to obtain high LA yields with moderate acid concentrations and at low temperatures from different substrates. All the substrates irrespective of their characteristics produced 60% or more LA molar yields with 20% H2SO4, 12% initial solids, and at 120oC. Molar yields obtained from different substrates are comparable or better than the literature reported yields on the basis of g-H2SO4/g-substrate. The primary difference between lignocellulosic substrates and soluble sugars is initial LA production rates. The initial LA production rates of lignocellulosic substrates are slower than the soluble sugars. Among the soluble sugars, fructose produced the highest LA yield (70.5%) and among the lignocellulosic substrates it was with Solka Floc (64.5%), whereas starch yielded 68% LA. Decrystallization of cellulose is an exothermic process because the hydrogen bonds are broken and the bond energy is released. Therefore, the temperature is expected to increase if there is decrystallization. Thus, this study focused on experimental verification on detecting decrystallization of cellulose by tracing the temperature in a system where pure crystalline cellulose is treated with sulfuric acid. Different substrates were used including, pure cellulose (Avicel PH 101), sulfuric acid treated pure cellulose (decrystallized), cellobiose, and glucose, to differentiate the extent of energy release by two different types of bonds. Decrystallization of cellulose occurs at room temperature with sulfuric acid up to 55-72%. Avicel crystallinity decreased substantially with increase in the acid concentration at room temperature and the decrease in crystallinity was from 67.9% to 4.2%. A sharp increase in temperature between Avicel-acid and Avicel-water could be observed within initial few seconds (15-25 seconds) with sulfuric acid up to 55-72%. However, the increase in temperature due to heat release was not entirely from decrystallization (hydrogen bond energy); a part of heat also came from hydrolysis of decrystallized cellulose to low DP cellulose and/or glucose. The heat releasing patterns and the crystallinity indices of 5 min and 30 min acid treated samples corroborate the fact that the higher the crystallinity, the higher the heat release. Initial enzymatic hydrolysis rates of 65-72% acid treated Avicel at room temperature were sharply increased, while rates were relatively sharper with 60% acid treated Avicel than with 55% acid treated and untreated Avicel. Data proves that decrystallization occurs at acid concentrations below 55%, if proper temperature is chosen. Avicel treated with 50% H2SO4 and at 90oC for 5 min released heat significantly and the loss in crystallinity was by 9%.