TREATMENT OF LIGNOCELLULOSIC BIOMASS WITH SURFACTANTS

CROSS-REFERENCE TO RELATED APPLICATION

This application claims the benefit of U.S. Provisional Patent Application Serial No. 60/926,662, filed April 27, 2007, the contents of which are hereby incorporated herein by reference.

BACKGROUND OF THE INVENTION

The present invention is directed methods and systems for enhancing the removal of lignin during a pretreatment process, and more particularly to the addition of one or more surfactants during pretreatment of a lignocellulosic biomass.

The US transportation sector is the largest contributor to greenhouse gas (GHG) emissions and is almost totally fueled by petroleum, about two thirds of which is imported with unstable regions holding most reserves. To address this situation, President Bush announced a goal of displacing 20% of fossil fuels by alternatives by 2017. Conversion of inexpensive and abundant cellulosic biomass to fuels has powerful attributes for production of sustainable transportation fuels. Producing ethanol from these materials is particularly well suited to the goals above due to the very high yields possible, the dramatic cost reductions that have been realized to make it competitive now, and the power of modern biotechnology for substantial additional cost reductions.

What is needed is the development of leading technologies for the integrated operations of pretreatment, enzymatic hydrolysis, and fermentation that dominate conversion costs to ethanol for existing biomass sources selected for their abundance, cost, and susceptibility to provide baseline performance data. In addition, advanced technologies would be developed to enhance yields while reducing costs. The results would provide a foundation for industry to commercialize cellulosic ethanol technology.

Cellulosic ethanol offers powerful benefits including near zero net greenhouse gas emissions, enhanced energy security and balance of trade, and creation of rural jobs. State-of-the-art technologies are needed for biologically converting abundant cellulosic biomass to ethanol and develop advanced processes that reduce costs while enhancing yields. Particular attention should be focused on integration of pretreatment, enzymatic

hydrolysis, and fermentation operations, the most costly process steps. To support process development, valuable data should be assembled in collaboration with experts on the availability, composition, and cost of biomass that is promising for ethanol production. The resulting feedstock and process information will facilitate assessments by industry of opportunities to competitively make ethanol from major fractions of agricultural residues and municipal solid waste and accelerate commercialization of processes for converting low cost cellulosic biomass into valuable products with major societal benefits.

Accordingly, what has been needed and heretofore unavailable is a system and method that overcome the deficiencies of existing processes for biologically converting cellulosic biomass to ethanol and advanced processes that reduce costs while enhancing yields. The present invention disclosed herein satisfies these and other needs.

SUMMARY OF THE INVENTION

For the effective enzymatic hydrolysis of lignocellulosic biomass, the lignin contained in them must be removed or modified by the pretreatment in order to avoid unproductive adsorptions. In the present invention, enzymatic digestibility of pretreated corn stover was enhanced by adding surfactants at 160-220° C. with and without acid. Compared with controls, most of the surfactants efficiently reduced the hydrophobicity on the corn stover surface. In addition, the adsorption of cellulases on pretreated samples were reduced due to the interaction of the surfactants with the lignin. Pre-treatment of the lignocellulosic biomass with surfactants also increased the digestibility of the lignocellulosic biomass and glucose recovery.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1 depicts the structure of a small piece of lignin polymer.

FIG. 2 is a schematic representation of hydrophilic lignin after treatment with a surfactant.

FIG. 3 shows data depicting the hydrophobicity of lignin after treatment with surfactants.

FIG. 4 is a schematic representation of the batch process of treating a lignocellulosic biomass with a surfactant.

FIG. 5 are photographs of batch reactors and their use in a heated sand bath.

FIG. 6 is a schematic representation of the a process using a flowthrough reactor for treating a lignocellulosic biomass with a surfactant.

FIG. 7 is a schematic representation of a flowthrough system.

FIG. 8 is a schematic of surface modification of a biomass with surfactants..

FIG. 9 is a block diagram of the effects of surfactants on pure lignin..

FIG. 10 is graph showing the glass transition temperature relative to different biomasses.

FIG. 11 is an infrared spectroscopy graph of different biomasses.

FIG. 12 shows data depicting lignin removal after treatment with surfactants in a batch reactor.

FIG. 13 shows data depicting lignin removal after treatment with surfactants in a flowthrough reactor.

FIG. 14 shows data depicting lignin removal after treatment with various surfactants.

FIG. 15 shows data depicting digestibility after treatment with various surfactants.

FIG. 16 shows data depicting the surgery recovery using TWEEN-80 and control samples of different biomasses.

FIG. 17 shows data depicting digestibility after treatment with a Tween 80 surfactant in a flowthrough reactor at different temperatures.

FIG. 18 shows data depicting digestibility after treatment with a Tween 80 surfactant compared to hydrolysis with BSA.

FIG. 19 shows data depicting glucose recovery after treatment with various surfactants in a batch system without acid pretreatment.

FIG. 20 shows data depicting glucose recovery after treatment with various surfactants in a batch system with one percent sulfuric acid pretreatment.

FIG. 21 is a flow chart of hydrolyzate recovery.

FIG. 22 is a set of method steps for hydrolyzate recovery.

FIG. 23 is a flow chart of treatment with blockers.

FIG. 24 is a set of method steps for treatment with blockers.

FIG. 23 is a flow chart of treatment with enzymes.

FIG. 26 is a set of method steps for treatment with enzymes.

DETAILED DESCRIPTION OF THE INVENTION

The method of the present invention is directed to the use surfactants during pretreatment of lignocellulosic biomass to remove lignin, to modify the biomass surface and to improve enzymatic hydrolysis with high yield, low cost. The method of the present invention includes pre-treating the biomass with one or more surfactants with or without an acid or a base at a temperature ranging from about 100° C. to about 210° C. Suitable surfactants include ionic and non-ionic surfactants that are suitable for use under high temperatures and are acid and/or alkali resistant in the mixture. The mixture may have a pH ranging from 1-14. Such suitable surfactants include, but are not limited to, Tween-80, Tween-20, PEG, DDBSA, glusopone/215, glusopone/225 and glusopone/625.

For the effective enzymatic hydrolysis of lignocellulosic biomass, the lignin contained in them must be removed or modified by the pretreatment in order to avoid unproductive adsorptions. In one aspect of the method of the present invention, enzymatic digestibility of pretreated corn stover is enhanced by adding surfactants at about 160° C. to about 220 ⁰ C, with and without acid. Compared with controls, most of the surfactants tested efficiently reduce the hydrophobicity on the corn stover surface. In addition, the non-specific adsorption of cellulases on pretreated samples can be reduced because surfactants are believed to interact with lignin.

Bioconversion of cellulosic biomass to fuels and commodity products has many potential benefits, and biomass refining provides a very promising alternative for efficient utilization of biomass. Furthermore, enzymatic hydrolysis of the cellulose in cellulosic materials to glucose followed by fermentation to ethanol is a very attractive route to produce sustainable liquid transportation fuels. However, enzymatic hydrolysis may be the most complex step in the bioconversion process due to the combination of substrate- related and enzyme-related effects. Thus, although the hydrolysis mechanism and the relationship between the structure and function of various cellulases have been extensively studied, the complex biomass structure confounds understanding the relative importance of these features and their roles, and reducing one barrier to digestion can enhance or disguise the importance of others.

It has been reported that enzymatic hydrolysis of bagasse was accelerated by pretreatment with 3 -33% wt nonionic surfactant (T ween 20; polyoxyethylene sorbitan monolaurate) at 170-190° C. The amount of the lignin remaining in the pretreated bagasse decreased by about 22-27% and the enzymatic hydrolysis rate increased, because of an increase in the surface area of cellulose accessible to enzyme, compared to those pretreated with water. This effect is based on the fact that surfactant makes hydrophobic degradation products extractable to water. The pretreatment effect varied with the HLB (hydrophile- lipophile balance) values of surfactant. The hydrophilic surfactant having high HLB values was useful for the extraction of hydrophobic degradation products from lignin and hemicellulose. In the pretreatment at 190° C, an addition of Tween 20 to the UCT-solvent (binary solvent with water, having upper critical-temperature in the manual solubility curve) increased the effect of UCT-solvent pretreatment. This is because the surfactant makes the separation of the degradation products easier.

It has also been reported that for the effective enzymatic hydrolysis of cellulosic materials, the lignin contained in them must be removed by the pretreatment in order to increase the surface area of cellulose for cellulase. Many pretreatment methods such as physical, chemical, physico-chemical and biological methods have been developed for separating lignin and hemicellulose in lignocellulose and for decreasing the degree of crystallinity of cellulose. It has been reported previously that the thermo-chemical pretreatment by UCT-solvent, having an upper critical-temperature (UCT) in the mutual solubility curve, is excellent as the pretreatment for bagasse and wood. In the case of

bagasse, about 76% of lignin was removed by pretreatment with a water-cyclohexanol mixture (named (UCT-solvent A; UCT-184° C.) at 200° C, and the hydrolysis rate was markedly accelerated. This pretreatment effect was caused by the formation of the homogeneous phase between water and solvent at a higher temperature than UCT, and from two actions taking place simultaneously: the degradation of lignin and the extraction of the degraded substances by cyclohexanol. Accordingly, it was concluded that surfactants may be good pretreatment reagents because, at adequate concentrations they have a high extraction ability for the degradation products through the formation of micelles.

Referring to FIG. 1, lignin is an obstructer during hydrolysis. Lignin is believed to affect enzymatic hydrolysis by unproductive binding with enzymes. Removing or modifying the lignin could be an effective way to reduce the enzyme loading. Lignin binds cellulose fibers together in a composite structure with excellent properties but also reduces the accessibility of cellulose to enzymes. Various studies have reported improved cellulose hydrolysis with increasing lignin removal, although differences have been reported in the degree of lignin removal needed. Lignin has been claimed to depolymerize and then repolymerize during hemicellulose hydrolysis, although no doubt in a different morphology that could change its impact on cellulose digestion. Generally, lignin modified by pretreatment is very important to enhance cellulose digestibility and that lignin removal provides even greater benefits.

Applying additives has shown promise in improving cellulase effectiveness, with most studies applying surfactants. In an early report, the nonionic surfactant Tween 80 enhanced the enzymatic hydrolysis rate of newspaper cellulose by 33%. Subsequent research found that several cationic surfactants improved performance with cellulose (Avicel) and tissue paper, while anionic surfactants did not. Biosurfactants, including sohorolipid and Tween 80, enhanced the saccharification rate of cellulose in Sigmacell 100 and steam exploded poplar by as much as a factor of 7 while decreasing enzyme adsorption on cellulose. Others showed that incubation of enzymes with surfactants was more effective than adding surfactants to substrates (filter paper) prior to hydrolysis. Tween 20 reduced thermal deactivation of cellulase and increased enzymatic cellulose and xylan conversion for lime pretreated corn stover by 42% and 40%, respectively, with loading on biomass found to be more important than the concentration in solution. Tween

80 (polyoxyethylene sorbitan monoleate) enhanced enzymatic hydrolysis yields for steam exploded poplar wood by 20% in the simultaneous saccharification and fermentation (SSF) process. Addition of Tween 20 reduced cellulase adsorption on solids and allowed a 50% reduction in cellulase loadings to obtain the same conversion for steam pretreated spruce (SPS) while having little effect on yields for delignified SPS.

As shown in FIGS. 2 and 3, surfactants will react with lignin, solubilizing lignin and modifying the biomass surface to decrease the hydrophicity of the substrate. Thus, addition of surfactants during pretreatment of biomass improves the lignin removal and reduces non-specific adsorption of cellulases to benefit the following enzymatic hydrolysis process. Previous research and debate on the effect of surfactants on enzymatic hydrolysis may be found in (1) Steve S. Helle, Sheldon J. B. Duff, "Effect of Surfactants on cellulose hydrolysis, 1993 (the hydrolysis of steam exploded wood was increased by 67% in the presence of sophorolipid); (2) M. Kurakake.H. Ooshima, "Pretreatment of Bagasse by Nonionic Surfactant for the Enzymatic Hydrolysis," 1994 (bagasse was pretreated with 3.33% nonionic surfactant (Tween 20) at 170- 190° C); (3) Malek Alkasrawi, Torny

Eriksson, "The effect of Tween-20 on simultaneous saccharification and fermentation of softwood to ethanol," 2003 (adding surfactants could decrease the enzyme loading and increase the ethanol yield during SSF of steam-pretreated wood); (4) Torny Eriksson, Johan Bόrjesson, "Mechanism of surfactant effect in enzymatic hydrolysis of lignocellulose," 2002 (discusses the mechanism of surfactants), the contents of which are each hereby incorporated herein by reference.

Compared to batch systems, flowthrough and countercurrent reactors have important potential advantages for pretreating cellulosic biomass including higher hemicellulose sugar yields, enhanced cellulose digestibility, and reduced chemical additions, but suffer from high water and energy use. Comparative data has been reported on xylan and lignin removal and enzymatic digestibility of cellulose for corn stover pretreated in batch and flowthrough reactors over a range of flow rates at 160 to 220° C. with only water and also with 0.1 weight percent sulfuric acid. B. Yang and C. Wyman, "Effect of Xylan and Lignin Removal by Batch and Flowthrough Pretreatment on the Enzymatic Digestibility of Corn Stover Cellulose," Biotechnology and Bioengineering, 86(l):88-98, 2004, the contents of which is hereby incorporated herein by reference. Increasing flow with just water enhanced the xylan dissolution rate, more than doubled

total lignin removal, and increased cellulose digestibility. Furthermore, adding dilute sulfuric acid increased the rate of xylan removal for both batch and flowthrough systems. Interestingly, adding acid also increased the lignin removal rate with flow, but less lignin was left in solution when acid was added in batch. Although the enzymatic hydrolysis of pretreated cellulose was related to xylan removal, as others have shown, the digestibility was much better for flowthrough compared to batch systems at the same extent of xylan removal. Cellulose digestibility for flowthrough reactors was related to lignin removal as well. The results suggest that altering lignin also affects the enzymatic digestibility of corn stover.

BATCH TUBE REACTORS

Referring now to FIGS. 4 and 5, to withstand the dilute acid concentrations, the batch reactors may be made of Hastelloy (C276) tubing. However, because Hastelloy caps are very expensive, Teflon plugs may be inserted in both ends that were then capped with stainless steel Swagelok fittings (Maine Valve and Fitting Co., Bangor, ME, USA). For example, the working volume of these tubular reactors may be about 13.4 mL. Thermocouple probes may be inserted along the centerline of the tubes and pushed to about 3/16" and 2" into the reactor. Reaction temperatures of 160, 180, 200, and 220° C. were applied at 0.0 and 0.1 wt% H ₂ SO ₄ with the solids concentration held at 5 wt%. Because temperature transients can impact biomass hydrolysis in reactor tubes, a three- bath procedure was developed to minimize these effects based on a thermal modeling approach. The sequence began by preheating each tube in boiling water for two minutes, followed by immediate transfer to a sand bath set at a temperature Tl selected through a modeling approach to minimize heat up times. After the tube is held for a specified time in this sand bath, it was transferred to a second sand bath set at the target reaction temperature T2. The time the tube was put into the second sand bath may be arbitrarily set as zero reaction time. After being subjected to the target temperature for a given time, the reactors are quickly transferred to an ice water bath to quench the reaction. Next, the tubes are removed from the water and dried, and the end caps and Teflon plugs were removed. The contents are then pushed out and separated into liquid and solid fractions by filtration for analysis.

FLOWTHROUGH REACTORS

Referring to FIGS. 6-9, flowthrough systems may be configured with a 1/2-in ID x 6-in length with an internal volume of 14.3 mL. These units may be constructed of 316 stainless steel parts using VCRTM (trademark of Swagelock Corporation) fittings, including one VCR male union (1/2-in), two gasket filters (316 stainless steel, average pore size 5 μm), two VCR glands (1/2-in x 1/2-in), two VCR nuts, and two VCR reducing fittings (1/2-inx 1/8-in). Such reactor parts may be obtained from Maine Valve and Fitting Co (Bangor, ME, USA). A 1/8-in stainless steel thermocouple (Omega Engineering Co., Stamford, CT, USA) may be installed at the outlet of the reactor to monitor temperature. Stainless steel tubing (316) may be used as a preheating coil (1/4-in OD x 0.35-in wall) and to connect the reactor with other system components as well the cooling coil (1/8-in OD x 0.028-in wall). The preheating coil is configured long enough to allow the incoming water to reach the desired temperature before it entered in the reactor, as measured experimentally. A high-pressure pump (Acuflow Series III Pumps, Fisher, Puerto Rico, USA) with a flow rate range from 0 to 40 mL/min, a pressure gauge (pressure range 0- 1500 psi, Cole-Parmer Instrument Co., Vernon Hill, IL, USA), and a back-pressure regulator (Maine Valve and Fitting Co., Bangor, ME, USA) may be used to control flow through the system.

To operate the flowthrough unit, about two grams of corn stover may be loaded into the reactor, which was then connected to the system. Distilled water at room temperature may be pumped through the reactor to purge air and then used to pressurize the system to a set pressure of 350 to 400 psig. The loaded biomass was completely wetted by this procedure. The preheating procedure may be the same as described herein for the batch tube reactor.

The enzymatic digestibility of cellulose by pretreatment of biomass may be enhanced with surfactants. For example, surfactants may modify the surface of pretreated biomass. As shown in FIGS. 10 and 11 the class transition temperature is relevant to the surfactant treatment of a biomass. A Parr reactor experimental apparatus was used for xylan depolymerization under the conditions of pH 1.4 (H ₂ SO ₄ ) at 160° C. for two minutes. As shown in FIGS. 12 and 13, various surfactants (FIG. 14) bind to lignin during the pretreatment process and improve the solubility of lignin thus improving lignin removal. As shown in FIGS. 15-17, adding surfactants during pretreatment also increase

the digestibility of the lignocellulosic biomass. As shown in FIGS. 18-19, adding surfactants during pretreatment also increase the glucose recovery of the lignocellulosic biomass in a batch process with or without pretreatment with sulfuric acid.

It is known that adding BSA during hydrolysis could help to prevent the unproductive binding of enzymes. B. Yang and C. Wyman, "BSA Treatment to Enhance Enzymatic Hydrolysis of Cellulose in Lignin Containing Substrates," Biotechnology and Bioengineering, 94(4): 611-617, 2006,, the contents of which is hereby incorporated herein by reference. Yang and Wyman reported that adding cellulase and bovine serum albumin (BSA) to Avicel cellulose and solids containing 56% cellulose and 28% lignin from dilute sulfuric acid pretreatment of corn stover. It was demonstrated that little BSA was adsorbed on Avicel cellulose, while pretreated corn stover solids adsorbed considerable amounts of this protein. It was also reported that cellulase was highly adsorbed on both substrates. Adding a 1% concentration of BSA to dilute acid pretreated corn stover prior to enzyme addition at 15 FPU/g cellulose enhanced filter paper activity in solution by about a factor of two and beta-glucosidase activity in solution by about a factor of 14. The results suggested that BSA treatment reduced adsorption of cellulase and particularly beta- glucosidase on lignin. Of particular note, BSA treatment of pretreated corn stover solids prior to enzymatic hydrolysis increased 72 hour glucose yields from about 82 to about 92% at a cellulase loading of 15 FPU/g cellulose or achieved about the same yield at a loading of 7.5 FPU/g cellulose. Similar improvements were also observed for enzymatic hydrolysis of AFEX pretreated corn stover and Douglas fir treated by SO ₂ steam explosion and for simultaneous saccharification and fermentation (SSF) of pretreated corn stover, hi addition, BSA treatment prior to hydrolysis reduced the need for beta-glucosidase supplementation of SSF. The results are consistent with non-specific competitive, irreversible adsorption of BSA on lignin and identify promising strategies to reduce enzyme requirements for cellulose hydrolysis. As shown in FIG. 20, lignocellulosic biomass pretreated with Tween-80 surfactant has almost the same digestibility as BSA blocking.

COUPLED PRETREATMENT AND ENZYMATIC CONVERSION TECHNOLOGIES

The pretreatment process described herein may be further developed by advanced technology that exploits combining washing feedstocks with room temperature dilute acid; supplementation of simultaneous saccharification and fermentation (SSF) with beta- glucosidase, xylanases, beta-xylosidase, and other enzymes; use of cellobiose fermenting yeast; and addition of surfactants or proteins to maximize total sugar recovery from the solids for the combined operations of pretreatment and enzymatic hydrolysis while minimizing total protein and acid demands.

The previous elements of the system and method described herein yield vital state- of-the-art processes for pretreatment, enzymatic hydrolysis, and fermentation, but costs must be reduced as much as possible, with higher yields with less enzyme the most powerful levers. In addition, eliminating or at least reducing acid use while maintaining or improving yields has significant value in both cost and impact on other operations. For example, a feedstock may be selected for initial development of lower cost processes, and then confirm performance with the others. The benefits of washing the feedstocks with dilute sulfuric acid at room temperature may be evaluated, since it has been shown that acid demands are reduced by avoiding the bisulfate shift that otherwise significantly reduces acid effectiveness. Materials with and without the acid wash may then be pretreated with very dilute (<0.5%) and no sulfuric acid at low severities (moderate temperatures and/or short reaction times) to limit hemicellulose sugar losses. Following identification of conditions to maximize total sugar recovery from pretreatment coupled with enzymatic hydrolysis with cellulase, one may apply xylanases, beta-xylosidase, and other hemicellulase enzymes with different loadings of cellulase and beta-glucosidase enzymes to pretreated solids to solubilize xylose and other hemicellulose components left after low severity pretreatments in addition to glucose based on recent results showing great promise in the reducing total enzyme protein loadings (and therefore costs) to achieve a particular yield. Yields of glucose, xylose, mannose, arabinose, and galactose may be plotted against total protein loadings to identify enzyme combinations that maximize total sugar recovery from the solids and from combined operations of pretreatment and enzymatic hydrolysis while minimizing total protein and acid demands.

Cellobiose fermenting organisms combined with additives will be developed to reduce enzyme loadings and improve performance, with focus on one feedstock followed by validation on others. Research may show that Brettanomyces custersii by itself or Brettanomyces clausenii in coculture with Saccharomyces cerevisiae achieved high SSF ethanol yields without beta-glucosidase supplementation, cutting protein costs. Recent work showed that adding surfactants used by others such as Tween 80 or non-enzymatic proteins improved performance, lowering enzyme costs. Accordingly, additives in combination with cellobiose fermenting yeast will be developed to eliminate beta- glucosidase supplementation and reduce cellulase loadings while maintaining or even improving yields. Hemicellulases may be used as a supplement to provide a systematic investigation for the first time of the benefits of using additives, cellobiose fermenting organisms, and hemicellulases to achieve high yields with much less protein use and more mild severity pretreatment that reduce costs.

FIG. 21 is a flow chart of hydrolyzate recovery. FIG. 22 is a set of method steps for hydrolyzate recovery. FIG. 23 is a flow chart of treatment with blockers. FIG. 24 is a set of method steps for treatment with blockers. FIG. 23 is a flow chart of treatment with enzymes. FIG. 26 is a set of method steps for treatment with enzymes.

Other features and advantages of the present invention will become more apparent from the following detailed description of the invention, when taken in conjunction with the accompanying exemplary drawings. Further modifications and improvements may additionally be made to the system and methods disclosed herein without departing from the scope of the present invention. Accordingly, it is not intended that the invention be limited by the embodiments disclosed herein.