Return to Table of Contents
Return to Table of Contents |
Biofuels Program |
Approach/Background: Our primary goal is reduce the cost of cellulases that act on pretreated biomass used for bioethanol production by increasing the specific activity of the enzymes in the cellulase complex. We have shown that EI functions with a high degree of synergism with fungal exoglucanases, especially T. reesei CBH I. We have also shown that the artificial ternary cellulase system, which consists of A. cellulolyticus EI, T. reesei CBH I, and T. fusca E3, can release 85% as much reducing sugar from crystalline cellulase as the native T. reesei ternary system (i.e., EI, CBH I, and CBH II). This result is encouraging for the ultimate success of engineered cellulase systems, because the artificial enzyme system described above was necessarily tested at 50oC, a temperature far below that considered optimal for one key component, EI. Our research plan for FY 1996 recommended the use of SDM to improve the activity of EI on biomass.
Throughout FY 1996 and FY 1997 efforts were also aimed at improving the basic cellulase assay. Although cellulase enzymes are widely sold, and their industrial use projected, on the basis of the filter-paper unit of activity, the traditional filter-paper assay is severely limited as a predictor of cellulase performance in the extensive (80%–90%-plus) saccharification of actual industrial lignocellulosic substrates. These limitations are traceable both to the chemical and physical differences between filter paper and the industrial substrates, and to the nonhomogeneous nature of most cellulosic substrates (filter paper included), which means that assays run to very limited extents of conversion (such as the 4% conversion target in the filter-paper assay) measure the digestibility of only the most easily digestible fraction of the substrate, and reveal little about the convertibility of the bulk of the substrate.
The fermentation performance of Zymomonas could be enhanced by transforming a wild-type Zymomonas strain with a gene coding for an effective cellobiase. The primary effort during FY 1997 was to acquire a suitable cellobiase gene either by screening NREL culture collections with RT-PCR or by materials transfer from external sources.
Characterization of the induction rules for the T. reesei enzyme complex was also an objective of FY 1997. The enzymatic saccharification of cellulosic biomass by the T. reesei enzyme complex is more effective when the microorganism is grown in the presence of the biomass substrate it will ultimately saccharify.
Status/Accomplishments: SDM Mutagenesis of Acidothermus EI. To meet the objective of the FY 1997 C milestone to “Improve the specific activity of the EI endoglucanase by 10% acting on pretreated biomass (yellow poplar),” we refined the rational design approach initiated in FY 1996 with the aid of a new 1.8-Å crystallographic structure of the enzyme obtained by Dr. Karplus at Cornell University. We have targeted modification of the enzyme that could improve the saccharification performance of the enzyme measured by DSA by systematically replacing two hydrophobic, surface amino acid residues thought to interact strongly with the cellulose substrate. We used PCR mutation to generate 26 mutant EI coding sequences and, after verifying the mutation sites by PCR DNA sequencing, employed microtiter plate assays to determine that 12 of these mutations yielded active enzymes. Ten-L cultures of these 12 transformed Escherichia coli that expressed active enzymes were grown and each mutant enzyme purified to homogeneity using a improved three-step column chromatographic method. The 12 purified EI endoglucanase enzymes (including the EI control) are undergoing rigorous kinetic analysis. The Y82 and W42 amino acid residues thought to interact with biomass have been probed with most of the replacement residues planned (i.e., Ala, Glu, Gln, and Arg), thus creating eight new interactive surface-modified EI mutants.
Acquisition of Cellobiase Gene for Zymomonas Transformation. USDA researchers recently identified the cellobiase produced by Candida peltata as an ideal enzyme for use in biomass conversion because of its high resistance to end product inhibition (i.e., cellobiose). NREL and USDA signed a Materials Transfer Agreement in November 1996 to permit NREL to work directly with the source microorganism. NREL successfully purified milligram quantities of the cellobiase from C. peltata culture supernatant with a new two-step chromatographic procedure. This enzyme was subjected to Edman sequencing; this peptide sequence will be used to design a series of DNA probes that will permit us to identify the fragment of native DNA that harbors the cellobiase coding sequence by DNA hybridization.
Improved Cellulase Assay. A new saccharification assay has been devised at NREL in which a continuously buffer-swept membrane reactor is used to remove the solubilized saccharification products. This DSA is a reliable predictor of the performance of combinations of cellulase and substrate under simulated SSF conditions but retains the analytically more direct nature of a saccharification reaction. Improvements in the apparatus and procedures for the DSA have resulted recently in significant improvements in the precision of the results. For assays that exhibit 96-h digestibilities in the vicinity of 45% saccharification, the standard deviation has been reduced from the 1.7% saccharification reported in February 1997 to less than 0.7%.
Summary Date: August 1997
Objectives: To help NREL researchers design protein engineering-based strategies capable of increasing the performance of A. cellulolyticus EI acting on pretreated biomass.
Approach/Background: The 2.4-Å resolution crystal structure generated during the past year has already proven invaluable in developing protein engineering strategies for the EI enzyme. During FY 1996, work at NREL to use protein engineering was initiated and focused on this enzyme. Because there was limited expertise initially at NREL in the areas of protein crystallography, protein structure, and cellulase mutation strategies, this consultation provided the team with an invaluable new asset.
Status/Accomplishments: Dr. Wilson participated in brainstorming sessions held both at Cornell University and at NREL to discuss strategies to improve EI. A special protein projection theater at Cornell was used to evaluate the structure of EI in a team setting, which permitted Cornell and NREL scientists to investigate aspects of the enzyme structure simultaneously. A decision was made to initiate mutation strategies by targeting amino acids that define the cellulose binding sites on the interactive surface of the enzyme, i.e., Tyr 82 and Trp 42. An active site residue, Trp 245, was also identified for mutation. Several amino acid residues were also targeted for replacement to improve the thermal tolerance of EI, should that be a desirable trait for bioprocessing.
Publications and Presentations:
Summary Date: August 1997
Objective: Develop high-quality x-ray crystallographic structures of important cellulases identified by NREL or NREL subcontractors.
Approach/Background: Genetically improving an enzymatic activity can best be approached in a directed fashion if a reliable three-dimensional crystal structure is known for the target protein. X-ray crystallography provides the only known approach to solving this problem for a protein the size of the EI catalytic domain.
Status/Accomplishments: Improvement in the quality of the EIcd structure. The initial structural analysis was done at 2.4-Å resolution using a rotating anode source. Unfrozen crystals diffract to 1.8 A resolution using the Cornell High Energy Synchrotron Source (CHESS), so attempts were made to improve the resolution of the liganded structure. To obtain data from EI crystals (which have one long axis) in a reasonable time at CHESS, the crystals must be frozen. Crystals were thus grown in the presence of both cellobiose and cryosolvent (20% ethylene glycol). Despite the additional care, frozen crystals diffracted to only 2.4-Å resolution with the synchrotron source. Small (0.1x 0.1x 0.3-mm) crystals obtained without cellobiose, but with cryosolvent (20% ethylene glycol), diffracted to 1.6-Å resolution. The data to 1.8-Å resolution from a frozen crystal have now been collected at CHESS. The number of observable data more than doubled from 47,386 at 2.4-Å resolution to 119,463 at 1.8-Å resolution. Because of the non-isomorphism, the structure was determined by molecular replacement method and has been refined to an R-factor of 0.18. The structure is high quality so it can now be analyzed for its bond-angle, bond-distance, and peptide-bond strains. The understanding of such strains will help us understand their importance in thermostability.
Crystallization of EIcd with cellodextrins. EIcd is an atypical catalytic domain (CD) that can hydrolyze filter paper without a cellulose binding domain (CBD). One hypothesis was that besides the four platforms that interact with cellulose, EI has additional platforms that allow it to interact with the crystalline cellulose more tightly. To obtain more extensive information about cellulose interactions with EI, attempts were made to crystallize EIcd in the presence of longer substrates such as cellotriose, cellotetraose, cellopentaose, and even carboxymethyl cellulose (CMC). In 240 trials, only small protein crystals of EIcd (0.1 x 0.1x 0.4 mm) were obtained in the presence of cellotetraose. The cellotetraose also crystallized alongside. The protein crystals, confirmed by a protein-specific dye, have similar crystal morphology with the EIcd:cellotetraose crystals, but were too small to be analyzable with our in-house source. Also, much of cellotetraose crystallized alongside, so its concentration on the enzyme could be too low, and the ligand is expected to have very low occupancy on the enzyme.
Summary Date: August 1997
Objective: To genetically modify potato plants to direct cellulase enzyme production to leafy (non-crop) parts, and preserve the natural state of the edible tuber (crop) portion. By targeting expression to unused, waste portions of plants, enzyme production adds to, rather than competes with, the value of the original crop.
Approach/Background: The potato crop was chosen for transformation because of PNNL’s current relationship with regional potato growers and processors. This relationship affords a potential commercial outlet for the technology proposed in this statement of work. In addition, potato plants typically transform efficiently using Agrobacterium-mediated methods. Also, potato may be quickly and clonally propagated with tuber explants. This type of asexual reproduction offers significant advantages in maintaining genetic characteristics. To speed up the progress of this work, tobacco will be employed as a model plant species for the bulk of the work in FY 1997. Tobacco presents a potentially viable opportunity in itself for use as a bioreactor for producing industrial enzymes. Target levels of active cellulase accumulation in leaves are ultimately greater than 5% of total soluble protein. This may prove a difficult goal to achieve, but is definitely within the realm of technical feasibility, as previously demonstrated by others. The following tasks will be completed with two distinct cellulase clones chosen among available endoglucanases, exoglucanases, and beta-glucosidases. Specifically, the target enzymes will be T. reesei CBH I and A. cellulolyticus EI endoglucanase. Initially, each of the two selected cellulase genes will be engineered for expression in plants and individually introduced into potato plants.
Status/Accomplishments: This work is ongoing. PNNL has acquired two rubisco small subunit genes (RbcS) from W. Grussam at U.C. Berkeley. The rubisco small subunit is a nuclear-encoded gene imported into chloroplasts after synthesis in the cytoplasm. It is one of the most abundant proteins in nature. Various signal sequences are also being collected and subcloned from various sources. Dr. Dai has constructed a series of chimeric CBHI genes with various signal peptides and 5-in. end untranslated leader sequences, which will be regulated by the rubisco RbcS-3C or Mac promoter (mannopine synthase promoter + enhanced region of 35S promoter. Without any signal peptide, the gene product will be targeted to the cytoplasm. With the rubisco Rbc-2A signal peptide, the gene product will be targeted to the chloroplast. With the pathogen resistance gene signal peptide, the gene product will be targeted to the apoplast. With the endoplasm reticulum signal peptide, the gene product will be targeted (or retained) in the endoplasm reticulum. These chimeric genes have been integrated into a binary vector that permits transfer into Agrobacterium tumefaciens. These constructs have also been transferred into tobacco leaf explants with A. tumefaciens infection. The infected leaf material is currently under shoot induction. Within approximately 1 month, enough shoot/leaf material will be produced for preliminary transformant screening.
Summary Date: August 1997
Objective: To develop new methods for measuring and classifying cellulase enzymes, especially endoglucanases. This work will contribute to a more efficient and timely execution of cellulase protein engineering efforts throughout the program.
Approach/Background: Endoglucanases from currently available sources require a high enzyme-to-substrate ratio for efficient hydrolysis of lignocellulolytic substrates. Some 50–80 endoglucanases have been reported in the literature. The search for improved sources of cellulolytic enzymes has been approached by screening native cellulase producers, and through the use of recombinant techniques. The latter method has been used to alter specific endoglucanases to produce novel enzymes with some novel properties; however, assessing the usefulness of the new enzymes is often difficult. Their mode of action has been poorly characterized, and there remains a need to compare the best suitable endoglucanases with internationally standard methodology. This work will develop a reliable assay method for cellulase enzymes based on the characterization of the change in molecular weight distribution of cellulose chains produced by the action of cellulases on high molecular weight substrates.
Status/Accomplishments: Carboxymethylcellulose (CMC), a derivatized and soluble cellulose, is generally used for endoglucanase kinetics. The corresponding reaction is followed either by release of reducing sugars (RS) or viscometrically. Both approaches are widely used, but they have several shortcomings. The methods based on detecting RS allow expression of enzymatic activity as the number of glycosidic bonds hydrolyzed per unit time during the initial period of hydrolysis, but do not indicate the point of breakage along the polymer chain. Viscometry measures hydrolysis according to internal bond breakage, but does not allow molecular weight to be directly measured. High-pressure size exclusion chromatography, in conjunction with multi-angle laser light scattering (HPSEC-MALLS), provides reliable molecular weight measurements of polysaccharides without calibration standards. The subcontractor used HPSEC-MALLS to directly measure the number average molecular weight (Mn) of CMC during hydrolysis by highly purified endoglucanases and related this key parameter to the number of glycosidic bonds broken (measured by RS concentration) and to the decrease in hydrolysate viscosity.
HPSEC-MALLS was used to follow the hydrolysis of EI and E5 from Thermonomonospora fusca and EG I and EG II from Trichoderma reesei. This technique provided the direct measurement of Mn of CMC, which was related to the decease in hydrolysate viscosity (measured by capillary viscometry) and to the number of glycosidic bonds broken (measured by the BCA method). The combination of these three methods revealed differences between the modes of action of the endoglucanases on CMC. Also, the linear equations have been established to convert viscosity to Mn of CMC at the initial period of hydrolysis by all endoglucanases.
Summary Date: August 1997
Objective: Cellulases with higher turnover numbers could dramatically reduce the cost of using these enzymes in bioethanol production, if they could function under adverse conditions such as high temperature, high salt concentrations, and extremes of pH that would favor increased bioconversion of cellulosic feedstocks. Of special interest would be the discovery of a cellulase that was stable at acidic pH and high temperature so the enzymes could be added directly to a dilute-acid treatment of biomass for simultaneous pretreatment and hydrolysis. Strategies to enhance the efficiency/specific activity of cellulase enzymes through the attachment of pentaammineruthenium (III) and other metal complexes will also be pursued.
Approach/Background: Microorganisms that grow under extreme conditions such as very low or high temperature, high salt, high pressure, or acidic or alkaline pH are called extremophiles, and the “extremozymes” they produce function under the same extreme conditions under which the microorganism grows. The isolation and utilization of extremozymes offers two advantages for bioprocessing: they can be used in process steps that require high temperature and caustic pH, and their intrinsic reaction rates can be higher that those of mesophiles. Currently, a number of sources of extremozymes are available from the collection of thermophilic, barophilic, and halophilic bacterial at ORNL. These include pure and mixed cultures obtained by the DOE Subsurface Drilling project, as well as Methanococcus jannaschii. There is evidence for the existence of cellulases from many of these organisms.
The rationale for this strategy is that ruthenium and other metal complexes (iron porphyrins and cobalt macrocycles) can possess oxidoreductase activity. Attaching these complexes to cellulases should confer on them oxidoreductase activity and, providing their native hydrolytic activity is unhindered to any great extent, may increase their efficiency toward lignocellulosic substrates if they can oxidize lignin as well as hydrolyze cellulose.
Status/Accomplishments: None due
Major Project Reports: None
Summary Date: August 1997
Objective: To develop enough information at the pilot scale to justify constructing a demonstration facility for converting biomass to ethanol.
Approach/Background: NREL and SWAN Biomass Company, a partnership between Amoco Ethanol Development Corporation and Stone and Webster Engineering Industrial Technology Corporation, have completed Phase 3 of the Amoco/NREL Cooperative Research and Development Agreement (CRADA) 91-0003. Phase 3 included a laboratory program to evaluate the yeast strains developed at Purdue University and support pilot scale testing, a series of runs in the NREL process development unit (PDU) using corn fiber as the feedstock, and process design and economic modeling.
Status/Accomplishments: Enormous progress was made in defining and verifying the biomass-to-ethanol process during Phase 3. SWAN’s pretreatment process produced substrate that was highly reactive with cellulase enzymes, with virtually all the hemicellulose converted to monomeric sugars, and without high concentrations of reversion inhibitors such as furfural. The Purdue yeast was able to ferment glucose and xylose simultaneously. Two PDU runs of about 1,000 hours each were completed, and more than 10 tons of solid residue were produced to be tested as an animal feed. Laboratory fermentations closely mimicked the results in the PDU, which confirmed that the work completed at smaller scale will translate easily to the larger unit. The data were successfully modeled, and the model was used to derive a commercial design and cost estimate. The estimated cost for ethanol produced from corn fiber in a commercial facility is about $0.86 per gallon, including a 15% internal rate of return on the capital.
Before Phase 4 is initiated, a number of technical problems need to be solved, and a fully integrated run that demonstrates the proposed commercial process configuration in the PDU must be made. Most of the technical problems can be addressed in tests with any of a number of feedstocks. The most significant unresolved generic problem is how to eliminate or control the negative impact of fermentation inhibitors on the rate of fermentation. Programs have been established outside of the CRADA to address this issue, and preliminary results indicate that a commercially acceptable solution to the problem has been found. For corn fiber feedstock, it is also critical to learn how to collect and dry the solid coproduct more efficiently.
Summary Date: August 1997
Objective: To identify and evaluate opportunities for bioethanol and to form partnerships with key players.
Approach/Background: Much of the western United States is faced with critical issues about using and managing its forest resources. In particular, suppressing forest fires has caused many small-diameter, dead, and diseased trees and underbrush to accumulate and create a severe fuel loading problem. Fires are so intense and difficult to control that they destroy everything in vast areas. Severe flooding and erosion often follow. Also, the unnatural ecosystem created by decades of fire suppression prevents the growth of healthy, high-quality trees and diminishes wildlife habitat. Biomass overstocking increases evapotranspiration and diminishes groundwater.
One solution is to thin these stands and recreate the original ecosystem. However, this would be expensive, and the trees are of low economic value. One solution is biomass conversion to fuel ethanol and cogenerated electricity. This option deals with many issues and could help satisfy the demand for fuel oxygenates such as ethanol and ETBE. There are also synergistic benefits to the biomass-electricity industry.
Two partnerships were formed, one each for California and Colorado, to evaluate the feasibility of producing ethanol from forest thinnings. The partners in the Northeastern California Ethanol Manufacturing Feasibility Study are NREL, the Quincy Library Group, California Energy Commission (CEC), California Institute for Agricultural Research, Plumas Corporation, and TSS Consultants. The Front Range Forest Health Partnership is spearheaded by NREL and the USDA Forest Service, and has many local and regional partners. This partnership will also evaluate waste from the urban forest. The Colorado Department of Public Health estimated that in 1995, 2.3 million cubic yards of yard wastes were disposed of in the state.
To complement these studies, NREL and Delta-T formed a CRADA to develop a process design and a rigorous capital and operating cost estimate for a commercial-scale biomass-to-ethanol facility. NREL is to provide a conceptual design and technology performance data for a biomass conversion technology, and Delta-T is to then apply its engineering expertise to produce a detailed engineering design and cost estimate.
NREL and the State of California are cosponsoring a Life Cycle Assessment (LCA) of ETBE production from California feedstock such as biomass from forest thinning and rice straw diverted from open-field burning. This partnership comprises NREL, California Environmental Protection Agency-Air Resources Board, CEC, the Resources Agency-Department of Forestry and Fire Protection, California Food and Agriculture Agency, TSS Consultants, and Ecobalance Inc.
Status/Accomplishments: Northeastern California Ethanol Manufacturing Feasibility Study. Tasks related to feedstock supply and delivery systems, site selection, and market and socioeconomic issues are complete. Tasks related to ethanol facility design and cost estimate, financial evaluation and sensitivity analysis, and environmental issues will be complete in September 1997. The feasibility study final report will be issued in the fall of 1997.
Front Range Forest Health Partnership. The initial information gathering is complete and manyof the data analyzed. Some plant design and economics work (including conclusions, barriers, and recommendations), remain before the draft report can be completed.
Delta-T CRADA. A softwood feedstock was chosen in response to growing concerns about forest health in the Western United States. The conversion technology chosen was a two-stage acid hydrolysis process for converting the six-carbon carbohydrates (primarily glucose) to readily fermentable sugars. A preliminary process flow sheet was completed, and capital and operating costs were estimated for the plant. A final detailed design and cost estimate are expected October 1997.
Life Cycle Assessment Partnership. A steering committee was formed to direct the study; its members are from the sponsoring entities and consultants to the study mentioned earler. A draft scoping document was prepared and will be sent to the stakeholders for review in September 1997. The LCA is expected to be completed by September 1998.
Summary Date: August 1997
Objective: Develop processes to convert lignins into hydrocarbon and oxygenated fuels compatible with reformulated gasoline (RFG).
Approach/Background: Lignin is a by-product of chemical pulping processes and is likely to become available as an unfermentable residue from lignocellulosic biomass ethanol production. Thus, the supply of lignin will progressively increase as ethanol plants that use straw, forest residues, etc., as feedstocks, are implemented. Lignin from ethanol plants is planned to be burned. An alternate strategy is to convert the lignin to fuels or other added-value products. The lignin would be depolymerized to simple monomeric and oligomeric O-containing aromatic units that would be used as intermediates for upgrading to higher-value products.
Since 1994, Dr. Shabtai's group at the University of Utah (U of U) has worked on the fundamentals of lignin depolymerization and upgrading to fuels and fuel additives compatible with RFG. The project has generated a significant number of interesting data. The Shabtai lignin process involves base catalyzed depolymerization (BCD) of lignin in the presence of an alcohol (methanol or ethanol). KOH has been the base mostly used by the Utah group, yet other bases such as NaOH are possible. The BCD step also involves recovery of the depolymerized lignin and base. In one approach, the BCD step is followed by upgrading via hydroprocessing (HPR). This step involves catalytic hydrodeoxygenation and hydrogenation. In the HPR step the BCD product is converted into a mixture of polyalkylated naphthenes and branched hydrocarbons that should be fully compatible with RFG. In another approach selective ring hydrogenation (SRH) is applied in which O functions are preserved while the aromaticity of the product is reduced. The SRH step may be preceded by etherification. This step should lead to alkyl cyclohexyl ethers whose performance is as good as or better than current commercial oxygenated fuel additives.
Early this year a team of staff from the U of U, Sandia National Laboratory (SNL), and NREL formed to further develop the Shabtai lignin process. The U of U group continues to focus on the fundamentals of lignin depolymerization and upgrading to fuels and fuel additives. SNL is studying the process scaleups, including unit operations such as base recovery. NREL supplies and chemically characterizes lignin feedstocks and products generated by the other two groups.
Status/Accomplishments: Systematic studies have been performed at the U of U on the base catalyzed depolymerization step, and looks at the effects of lignin moisture content, reaction time, temperature, and pressure, base type and concentration, and the amount of methanol used. High lignin conversions (75%–100%) are obtained at temperatures (250·–300·C) that are above the critical temperature of the alcohols. Reaction times as short as 15 minutes (at 290·C) appear to result in complete lignin conversion. Water in amounts equivalent to 50%, and even 100%, of the lignin feed appear to cause only a slight decrease in BCD product yields. The optimum performance appears to lie in the direction of lower methanol concentration that results in lower solvent incorporation, base at about 75% of the lignin concentration, and reaction times toward 15 minutes at 270·C. NaOH works as well as KOH, but CaOH2 does not.
A sample of a hydrogenated BCD product has been analyzed at NREL. Its structure is consistent with those proposed for the BCD products from which it was produced. It is almost completely deoxygenated and appears to be a mixture of cyclohexanes (48%) and cyclohexenes (44%), with trace amounts of aromatic components (8%). There is evidence for incorporating the BCD reaction solvent, methanol, in the structure with addition of 2-3 methyl groups, appearing on both the ring and side-chain. Most of the product is in the range of monomers to trimers, with the extent of monomers estimated at 30%–50%. This level of monomers is partly due to the lack of hydrocracking activity in the hydrotreating catalyst used, and partly to the structure of the starting lignin that was a heavily modified commercial lignin and contained many carbon-carbon bonds formed by condensation reactions during its isolation. Lignins from the NREL ethanol process are not expected to contain these resistant bonds. Catalysts with higher hydrocracking activities are also being tested, so that higher levels of monomers should be obtained in the future.
SNL has begun work on the BCD reaction. It has replicated the work performed at the U of U and sent five products to NREL for characterization. Analysis of these samples has shown that they are chemically very similar to those produced at the U of U. A further set of samples have been prepared at SNL and sent to NREL for analysis to examine the reproducibility of this result. In addition, a set of 10 more samples have been sent to NREL for analysis. Some were produced at very short reaction times (<1 minute after reaching the reaction temperature) and will allow us to assess the speed of the BCD reaction.
Summary Date: September 1997
Objective: Develop a rapid method for measuring Zymomonas mobilis cell mass concentration and cell viability, or metabolic activity, in fermentation broths and biomass hydrolyzates. Ultimately, the objective is to extend the method to quantitation of Zymomonas cell mass concentration and cell viability in the presence of solids, i.e., such as would be encountered in simultaneous saccharification and cofermentation process samples. This effort is significant because current plating-based methods are slow and highly variable. The development of an effective method for quantifying cell mass concentration and cell viability in bioethanol process samples will greatly facilitate process development efforts by strengthening toxicity testing and process diagnostic capabilities.
Approach/Background: Traditional plate countand slide count-based methods for enumerating cell mass concentration are time consuming and will not provide correct toxicity information if the agar or slide surface environment alters the sample. For example, hydrolyzate toxicity testing can be skewed if hydrolyzate diffusion into the agar reduces the apparent toxicity (i.e., inhibition experienced by microorganisms at the surface of the plate). In general, these methods require substantial dilution of the sample. In many instances, the sampled microorganisms must be propagated in a nutrient-rich environment to be enumerated. Some cells will likely be resuscitated by such methods. To overcome the limitations of traditional approaches, we are pursuing the development of a dye-based spectrophotometric method. Dye-based methods are generally quicker than traditional methods because they do not require cell replication. Generally, incubation times of less than a few hours are sufficient with dyes; incubation times of several days are common with traditional plate methods. Most importantly, because incubation times are short in dye-based methods, cells can remain in the sample environment and thus the test can provide highly relevant diagnostic information.
Status/Accomplishments: This project is being carried out at NREL on a part time basis by N. Dowe, an employee of the Colorado Bioprocessing Center. The project has only begun and is proceeding well, but there are no key accomplishments to report yet. The ethanol process development team is working closely with the principal investigator to provide Zymononas cultures and ensure that anticipated interferences in process hydrolyzates are considered early on in the dye evaluation process. Candidate dyes will be evaluated for their ability to stain Zymomonas and discriminate between viable and nonviable (or metabolically active and metabolically inactive) bacterial cells. Dyes with good cell staining characteristics will be further assessed under a range of anticipated bioprocessing conditions.
Summary Date: August 1997
Objective: Develop and demonstrate integrated biomass-to-ethanol process technology that incorporates dilute-acid pretreatment and NREL's recombinant xylose-fermenting Zymomonas mobilis with hardwood sawdust as a model feedstock. Research will be guided by process engineering sensitivity analyses results that identify the process variables with the highest impact on overall ethanol production cost. The objective is to demonstrate integrated process technology that meets MYTP year 2000 performance goals, ultimately at scales that enable carbon mass balance closures to be assessed and thereby permit the integrity of performance data to be verified. This effort is significant because it establishes the performance level for first-generation enzymatic hydrolysis-based process technology and identifies specific unit operations or integration issues for which additional performance improvements have the potential to further reduce the projected bioethanol production cost.
Approach/Background: Early work will involve evaluating the main unit operations of the process— pretreatment, conditioning (detoxification), cellulase production and saccharification/cofermentation—and establishing performance targets for a base-case integrated process. Once the individual unit operations are sufficiently developed, work will focus on integrating the separate unit operations. In parallel, work to further improve selected specific unit operations will also be performed, particularly for areas where cost sensitivities are high. Initially, work will be carried out at the bench scale, i.e., the smallest scales possible. After initial process integration at the bench scale, work to demonstrate the process at a modestly larger scale, i.e., the minipilot scale, will occur. Carbon mass balance closures can only be reliably assessed at or above the minipilot scale.
Status/Accomplishments: FY 1996 work activities focused primarily on developing an effective hydrolyzate conditioning process (many were investigated) that enabled pretreatment and simultaneous saccharification and cofermentation (SSCF) processing steps to be effectively integrated. In addition, a xylose-fermenting Zymomonas strain was adapted to better tolerate inhibitory hydrolyzates. These developments culminated in the initial demonstration at the bench scale (0.4 L working volume), of an integrated SSCF process that achieved an ethanol yield of 54% at a cellulase enzyme loading of 12 FPU/g cellulose. FY 1997 work is building on these results. Initial work focused on demonstrating continuous ion exchange-based detoxification at the pilot scale. Ongoing work is directed at demonstrating integrated SSCF processing at the minipilot scale (above 10 L working volume). In addition, cellulase production research was revived in FY 1997 to develop base-case process technology for cellulase production using detoxified hydrolyzates and pretreated solids.
Summary Date: August 1997
Objective: Sensitivity analysis of the base-case biomass-to-ethanol process indicates substantial savings in capital and operating costs can be gained by using a cofermentation process in which hexose and pentose sugars are simultaneously fermented to ethanol in a single unit operation. Previously we developed a unique strain of the ethanologenic bacaterium Zymomonas mobilis that can simultaneously coferment the glucose and xylose prominent in many lignocellulosic feedstocks. The objective of this project is to develop improved Zymomonas strains for rapid and efficient cofermentation of glucose, xylose, and arabinose.
Approach/Background: Several microorganisms can efficiently ferment the glucose component in cellulose to ethanol; however, because there is no suitable biocatalyst, converting the pentose sugars such as xylose and arabinose in the hemicellulose fraction is more difficult.
Simultaneously cofermenting these pentose sugars is further hindered by the repressive effect of the glucose liberated during enzymatic hydrolysis of cellulose. A comprehensive survey identified Z. mobilis and Lactobacillus as promising microorganisms for further development as cofermentation biocatalysts.
The xylose assimilation and pentose phosphate genes have previously been cloned into Zymomonas from Escherichia coli. We have also cloned the arabinose assimilation and pentose phosphate genes into another Zymomonas host. We will now attempt to combine these three sets of genes into a single Zymomonas host strain.
Status/Accomplishments: A new strain of Z. mobilis has been metabolically engineered to simultaneously coferment the glucose, xylose and arabinose—prominent in many lignocellulosic feedstocks—to ethanol. Engineered strains that demonstrate the best cofermentation performance in sawdust hydrolysate will be identified for scaleup to the PDU. This new biocatalyst also appears to be ideal for fermenting the arabinose commonly found in agricultural residues such as corn fiber and in herbaceous energy crops such as switchgrass.
Seven genes, which encode the enzymes needed to convert xylose and arabinose to common intermediates of Zymomonas' central glycolytic pathway, were introduced, via a single plasmid, into Zymomonas under the control of strong promoters that direct their expression, even in the presence of glucose. The newly engineered strain can grow on and ferment xylose, arabinose, or glucose as a sole carbon source to ethanol and simultaneously ferment a mixture of 1% glucose, 2% xylose and 2% arabinose to ethanol at about 90% of the theoretical yield at 30·C without pH control. Further fermentation performance characterization with various pH, temperatures, and high sugar concentrations, will be evaluated in the future.
Summary Date: September 1997
Objective: To improve the transport of pentose sugars in xylose-fermenting strains of Zymomonas mobilis. This information is essential for developing superior biocatalysts to simultaneously coferment the hexose and pentose sugars in lignocellulosic feedstocks to ethanol.
Approach/Background: Sensitivity analysis of the base-case biomass to ethanol process indicates that there are substantial savings in capital and operating costs to be gained by use of a cofermentation process in which hexose and pentose sugars are simultaneously fermented to ethanol in a single unit operation.
Recently, NREL scientists metabolically engineered a new strain of the bacterium Zymomonas mobilis that can simultaneously coferment the glucose, xylose, and arabinose components prominent in many lignocellulosic feedstocks to ethanol. Research results suggest that the pentose sugar transport in this microorganism could be improved to maximize the sugar utilization. This is essential to further develop superior biocatalysts and processes to simultaneously coferment these sugars to ethanol.
Status/Accomplishments: The research was conducted to improve the xylose transport in Zymomonas. The native Zymomonas transporter GLF has a relatively low affinity toward xylose. One approach is to mutate the transporter to alter the kinetics of xylose transport.
Alternatively, the Escherichia coli xylose transporter xylE (with relatively high affinity to xylose) will be expressed in Zymomonas to determine whether the xylose/proton symporter is functional in Zymomonas and has better xylose transport capability.
Research was also conducted to determine the kinetics of arabinose transport in wild-type and metabolically engineered strains of Z. mobilis in the presence and absence of glucose or xylose. The results indicated that the affinity of GLF for arabinose is very low (lower than that of xylose), but the velocity appears to be fairly high. Research will be conducted to improve arabinose transport by mutagenesis of GLF.
Publications and Presentations: Annual Technical Report
Summary Date: September 1997
Objective: To develop an improved xylose-fermenting yeast through genetic engineering.
Approach/Background: The efficient fermentation of the xylose component in lignocellulosic hydrolysates remains a key economic bottleneck in a biomass-to-ethanol process. This work is directed toward increasing the ethanol yield and productivity of xylose-fermenting yeasts by overexpressing and deleting selected genes, evaluating their xylose fermentation performance, and implementing strategies that maximize anaerobic fermentation yield.
Status/Accomplishments: To better characterize oxygen regulation, two genes for pyruvate decarboxylase were cloned and sequenced from Pichia stipitis and their expression were studied.
Two alcohol dehydrogenases of P. stipitis were evaluated on various carbon sources and under oxygen-limiting conditions. The results showed that P. stipitis ADH2 is induced by oxygen limitation but not by glucose and that P. stipitis ADH1 is induced only under fully aerobic conditions when growing on a nonfermentable substrate.
Fermentative capability of P. stipitis overexpressing both XYL1 and XYL2 is greatly elevated. A punative clone of xylulokinase (XYL3) of P. stipitis was obtained.
Summary Date: September 1997
Objective: The goal of this project is to further develop Zymomonas as a microbial biocatalyst that effectively converts a variety of feedstocks to ethanol. This bacterium's high conversion yield, fermentation selectivity, and ethanol tolerance levels are key attributes for a commercially viable ethanol-producing strain. The narrow substrate utilization range of Zymomonas has largely been addressed in our breakthrough metabolic engineering work in 1994 for xylose utilization and 1996 work for arabinose utilization. Our FY 1997 work revolves around the characterization and further improvement of Zymomonas strains.
Approach/Background: The single Zymomonas strain that can grow on xylose and arabinose built toward the end of FY 1996 needed to be further characterized. There was a strong possibility it would need improvements in either performance stability, ethanol tolerance, or metabolic burden to the host cell. Detailed physiological and biochemical characterizations were undertaken to determine the capabilities of this new strain.
To address the stability issues, we will build on our ongoing work in which we are attempting to integrate a transposon that contains the tetracycline resistance gene into the Zymomonas chromosome. By locating the pentose phosphate and xylose assimilation genes within this transposon, we hope to introduce these proficiencies into Zymomonas in a stable format.
We have recently obtained a variety of non-genetically engineered strains of Zymomonas that are ostensibly hydrolysate, acid, thermo, or ethanol tolerant. Many or all of these qualities would be desirable in a commercial environment. These strains may not be as “hardy” as their parents (particularly after the plasmids are introduced), so we plan to evaluate and further characterize them for their robustness and fermentation capabilities.
Status/Accomplishments: A new strain of Z. mobilis was metabolically engineered to simultaneously coferment the glucose, xylose, and arabinose prominent in many lignocellulosic feedstocks to ethanol.
The plasmid pZB301 contains the seven genes required for Zymomonas to effectively ferment a mixture of 1% glucose, 2% xylose, and 2% arabinose to ethanol at about 90% of the theoretical yield at 30·C without pH control. Further fermentation experiments under various pH, temperataures, and high sugar concentrations have further defined the capabilities of this strain.
Twelve Zymomonas strains were evaluated as potential candidates for introducing our new pentose-plasmid constructs. Three hosts strains could effectively grow in and ferment glucose at concentrations as high as 25%. Furthermore, they appear to be more ethanol and pH tolerant than our current host strain.
We are pursuing a two-pronged approach to enhance the genetic stability of our recombinant Zymomonas in the absence of antibiotic selection. In addition to exploring a chromosome integration system based on an engineered mini-transposon Tn5, we started to investigate possible target DNA sequences for chromosome integration based on homologous recombination. The development of multiple integration systems in Zymomonas will allow us to not only perform multiple insertions, but also to inactivate unwanted genes such as those involved in by-product formation.
Publications and Presentations:
Summary Date: September 1997
Objective: To improve the transport of pentose sugars in xylose-fermenting strains of Zymomonas mobilis. This information is essential for developing superior biocatalysts to simultaneously coferment the hexose and pentose sugars in lignocellulosic feedstocks to ethanol.
Approach/Background: Sensitivity analysis of the base-case biomass to ethanol process indicates that there are substantial savings in capital and operating costs to be gained by use of a cofermentation process in which hexose and pentose sugars are simultaneously fermented to ethanol in a single unit operation.
Recently, NREL scientists metabolically engineered a new strain of the bacterium Zymomonas mobilis that can simultaneously coferment the glucose, xylose, and arabinose components prominent in many lignocellulosic feedstocks to ethanol. Research results suggest that the pentose sugar transport in this microorganism could be improved to maximize the sugar utilization. This is essential to further develop superior biocatalysts and processes to simultaneously coferment these sugars to ethanol.
Status/Accomplishments: The research was conducted to improve the xylose transport in Zymomonas. The native Zymomonas transporter GLF has a relatively low affinity toward xylose. One approach is to mutate the transporter to alter the kinetics of xylose transport.
Alternatively, the Escherichia coli xylose transporter xylE (with relatively high affinity to xylose) will be expressed in Zymomonas to determine whether the xylose/proton symporter is functional in Zymomonas and has better xylose transport capability.
Research was also conducted to determine the kinetics of arabinose transport in wild-type and metabolically engineered strains of Z. mobilis in the presence and absence of glucose or xylose. The results indicated that the affinity of GLF for arabinose is very low (lower than that of xylose), but the velocity appears to be fairly high. Research will be conducted to improve arabinose transport by mutagenesis of GLF.
Publications and Presentations: Annual Technical Report
Summary Date: September 1997
Objective: To develop an improved xylose-fermenting yeast through genetic engineering.
Approach/Background: Research is directed toward the complete development of genetic systems for Pichia stipitis, to determine the key genes for fermentation and respiration, the overexpression those genes required for efficient fermentation, and to disrupt or down-regulate those required for respiration.
Status/Accomplishments: A double auxotrophic mutant of P. stipitis was created for use as the host for advanced genetic manipulations in this microorganism. This mutant is useful for disrupting genes and the overexpressing several genes simultaneously.
Introducing the Saccharomyces cerevisiae dihydro-orotate dehydrogenase (DHODase) into P. stipitis enabled anaerobic growth on glucose medium within 72 hours. A 40-fold increase in cell density and neartheoretical ethanol yields were observed. Subsequent transfers resulted in significantly increasing the growth and fermentation rates. This is first demonstration of anaerobic growth of P. stipitis. However, the same strain did not grow anaerobically on xylose. This result implies other limiting factors in anaerobic xylose fermentation.
Several fermentative enzymes have been identified that affect ethanol production. Research is under way to maximize the expression of selected genes for ethanol production.
Summary Date: September 1997
Objective: To develop superior Lactobacillus biocatalysts to rapidly and efficiently coferment the pentose and hexose sugars in lignocellulosic feedstocks.
Approach/Background: Sensitivity analysis of the base-case biomass-to-ethanol process indicates that there are substantial savings in capital and operating costs associated with a cofermentation process in which hexose and pentose sugars are simultaneously fermented to ethanol in a single unit operation.
A comprehensive survey identified the bacterium Lactobacillus as a promising microorganism for further development as cofermentation biocatalysts.
In a first step to develop a thermotolerant ethanologenic cofermentation biocatalyst, a strain of Lactobacillus with superior resistance to dilute-acid hydrolysates at elevated temperatures was chosen. This strain also has the ability to utilize many sugars commonly found in lignocellulosic feedstocks, including glucose, cellobiose, mannose, and arabinose (but not xylose). Our earlier efforts have been to metabolically engineer this strain to produce lactate from xylose at near-theoretical yield.
Status/Accomplishments: Research was resumed in July 1997 with the hire of new staff to further metabolically engineer this potential superior ethanologenic cofermentation biocatalyst for ethanol production. Ethanol production pathway genes are currently being introduced into our genetically engineered xylose-fermenting Lactobacillus strain. This will allow us to redirect the carbon flow in this microorganism from pyruvate to ethanol. Inactivation of the native lactate dehydrogenase may be necessary to eliminate the lactate production to maximize ethanol yield.
Research will continue to optimize this recombinant Lactobacillus for simultaneous cofermentation of hexose and pentose sugars to ethanol.
Publications and Presentations: None
Summary Date: September 1997
Objective: Generate fundamental information on pH homeostasis in Zymomonas and its general stress protection system. Use this information to design research strategies to produce a more acid-tolerant and robust Zymomonas production strain.
Approach/Background: Commercially viable, large-scale production of ethanol from lignocellulosic materials requires a robust microorganism that can withstand a variety of adverse environmental conditions (temperature, acid, hydrolysate toxins, high sugar and ethanol concentrations, etc.) .
Many of the pretreatment procedures that NREL is exploring in its biomass-to-ethanol process yield sugar feedstreams that contain a variety of acids (primarily sulfuric or acetic). These acids are difficult and expensive to remove and physiogically impair most fermentative microorganisms. To improve biomass-to-ethanol process economics, microorganilsms tolerant to low pH environments should be developed.
Zymomonas is quite adept at growing in high sugars concentrations in mildly acidic environments (pH 5–6). Whereas the wild-type strain can grow at lower pH, NREL's metabolically engineered strains have great difficulty below pH 5, particularly when fermenting pentose sugars. A variety of ill-defined (for Zymomonas) cell protection or stress systems are activated that allow Zymomonas to survive in these low pH environments. These systems will be investigated in Zymomonas to determine viable approaches to enhance their acid tolerance.
Status/Accomplishments: Several genera of gram-negative bacteria employ a common protection system referred to as the general stress protection system (GSP). Zymomonas strains CP4 and 39676 were examined for the presence of genotypic and phenotypic characteristics of this system. A variety of studies that use hybridization probes from several gram-negative bacteria that contain GPS indicate that Zymomonas lacks such a system.
Lysine and arginine decarboxylases are important for maintaining pH homeostasis in many bacteria. Examination of our Zymomonas strains did not indicate any decarboxylase activity. These results indicate that the pH and ethanol tolerance of Zymomonas is most probably due to passive (i.e., membrane) protection systems.
Future work will involve introducing the Escherichia coli chaperone and DNA-binding proteins in Zymomonas to observe their effects on acid tolerance. Other relevant homologous or heterologous stress proteins will be overexpressed in Zymomonas. We hope they will boost its stress tolerance.
Publications and Presentations: Two bi-monthly reports
Summary date: September 1997
Objective: Develop and characterize strains of Clostridium thermocellum and Clostridium thermosaccharolyticum that have ethanol selectivity similar to more conventional ethanol-producing microorganisms. Document the maximum concentration of ethanol that can be produced by these thermophiles.
Approach/Background: Moderate thermophiles (optimal growth temperatures of ~60·C) have frequently been proposed for ethanol production, and are among the most frequently considered microorganisms for consolidated bioprocessing (CBP).
C. thermocellum is a cellulolytic, thermophillic, ethanol-producing strain that can excrete large quantities of cellulases. C. thermosaccharolyticum does not produce cellulases, but can ferment xylose and other sugars to ethanol. A co-culture of these microbes could lead to the development of a CBP for ethanol production.
Genetic systems for Clostridia species are not well developed. Substantial classical genetic approaches have been successful, but the cloning and expression of Clostridia genes has lagged behind because of the lack of suitable markers, vectors, and efficient transformation systems. This project will focus on developing such systems.
Status/Accomplishments: Optimization of electrotransformation has focused on both C. thermocellum and C. thermosaccharolyticum using MgCl2 and polyethylene glycol. Transformation frequencies of >104 per electroporation event are sought. This number was selected because it should be sufficient to allow recovery of homologous recombinants that will be introduced on nonreplicative plasmids during subsequent parts of this study.
The catabolic genes ack (acetate kinase) and hyd (hydrogenase) were chosen because of their known catabolic activity that diverts carbon flux away from ethanol and toward acetate. Knocking them out should increase the selectivity of the microorganism toward ethanol. PCR fragments for these genes have been produced from C. thermocellum, and library screening to identify specific clones that contain these genes is ongoing.
Publications and Presentations: Two bi-monthly reports
Summary date: September 1997
Objective: Establish chromosome integration system(s) for transferring foreign genes into Zymomonasmobilis and enabling them to be stably maintained and expressed.
Approach/Background: The current plasmid-bearing, xylose-fermenting Zymomonas strain is stable for batch simultaneous saccharification cofermentation (SSCF) processes in which the soluble xylose plays a role for selection pressure. However, we anticipate that stability of the plasmid could be problematic in an integrated continuous process and in high soluble glucose feedstreams without adding tetracycline for selection. Using antibiotics (such as tetracycline) in a commercial scale process for a biomass-to-ethanol plant is undesirable and uneconomical. Therefore, stable microorganisms must be developed without antibiotic selection.
To integrate genes into the Zymomonas chromosome, methodologies for chromosome integration in Zymomonas must first be established. One approach is to develop integration methods based on homologous recombination.
Status/Accomplishments: This research is conducted to develop integration methods based on homologous recombination. Several suicide plasmids that contain an antibiotic marker within Zymomonas tryptophan genes were constructed to facilitate chromosome integration.
After establishing an integration system in Zymomonas and characterizing the system, the integration of individual or a combination of xylose-fermenting genes into the chromosome can commence. Enzymatic activites of these genes need to be determined and sufficient gene copy numbers established. Eventually, all the pentose phosphate and xylose assimilation genes need to be integrated.
Publications and Presentations: Three monthly reports
Summary date: September 1997
Objective: To provide laboratory and pilot plant testing of a rice straw-to-ethanol conversion process using the proprietary technology of SWAN Biomass Company.
Approach/Background: The overall scope of the Gridley Project was to investigate the feasibility of converting rice straw to ethanol. The proposed conversion process used the technology developed by SWAN Biomass Company. The goal of Phase I of the NREL work was to provide laboratory and limited pilot plant testing of the SWAN process and to provide the data to SWAN for technoeconomic evaluation of the process. Phase II of the work was to demonstrate the process in NREL's Process Development Unit (PDU) and generate solid product for further testing.
Status/Accomplishments: During Phase I of the NREL work, pretreatment performance data with rice straw were generated using the proprietary Amoco Pretreatment Reactor (APR). In addition, data on fermentation performance were generated in the laboratory with bench-scale fermentors and the Purdue recombinant yeast. These data were supplied to SWAN. Phase II work has not yet commenced.
Publications and Presentations: None
Summary Date: August 1997
Objective: To receive, store, and shred lignocellulosic feedstock and ship to NREL as required for PDU operation. The subcontractor will also knife mill lignocellulosic feedstock and ship to NREL as required for other programmatic activities.
Approach/Background: When operating, the PDU converts 1 dry ton per day of feedstock to ethanol, but it does not have enough space and equipment to handle large quantities of feedstock. This subcontract provides the intermediate staging steps necessary to ensure continuous feedstock delivery to the PDU. The subcontractor receives and stores feedstock, then shreds it (if required) and delivers it to the PDU at the required rate.
Status/Accomplishments: The subcontractor has procured a warehouse facility for receiving and storing feedstocks. The shredder and knife mill have been installed and are operated by the subcontractor's personnel as required to meet NREL's programmatic needs.
Publications and Presentations: None
Summary Date: August 1997
Objective: To conduct detailed compositional analyses of dilute-acid pretreatment hydrolyzate liquors, both before and after various conditioning steps, in conjunction with selected fermentation toxicity testing, to identify the inhibitory compounds that adversely affect fermentation performance. This effort is significant because it is expected to generate information that will help us develop or improve strategies, for alleviating inhibition of fermentation performance in the presence of inhibitory biomass hydrolyzates.
Approach/Background: Relatively poor fermentative conversion of neutralized and overlimed dilute-acid pretreated hardwood feedstocks is attributed to inhibitory compounds that deleteriously affect the fermentative microorganisms used to ferment biomass sugars to ethanol. Suspected inhibitors include compounds in raw biomass and those formed or released during pretreatment. It is critical to the development of an economical biomass-to-ethanol process to thoroughly characterize the composition of dilute-acid pretreated biomass feedstocks and to understand the mechanisms by which various detoxification methods act to reduce toxicity. Rationale conditioning unit operations can be developed only when inhibitory compounds are identified and their probable fate(s) in an integrated biomass-to-ethanol process is understood.
Status/Accomplishments: This effort represents the continuation of the first year's effort. HPLC, GC, GC/ MS and other analytical methods continue to be used to identify and quantify suspected inhibitory compounds in dilute-acid pretreated materials. In FY 1996, acetic acid was shown to be the dominant inhibitory component in hydrolyzates. Although no other individual components were highly inhibitory, neutralized and overlimed hydrolyzates consistently show considerably greater toxicity than can be accounted for by acetic acid alone. Thus, additional work is currently under way to develop a better understanding of hydrolyzate toxicity and the mechanisms by which various conditioning and detoxification methods act. All work is being guided by NREL's ethanol process development team so that key results can quickly be incorporated into the base-case ion exchange-overliming conditioning and detoxification process. Outyear efforts will focus on assessing the efficacy of the base-case conditioning method for detoxifying feedstocks other than hardwood (switchgrass, agricultural residues, softwoods, etc.).
Summary Date: August 1997
Objective: Establish nutrient requirements for seed (process inoculum) production for cofermentating glucose and xylose in synthetic biomass hydrolyzates and characterize seed production in continuous culture under a range of operating conditions. Quantitatively understanding how seed production performance varies as a function of nutrient requirements and acetic acid level, as well as the potential for producing seed continuously, will enable strategies for minimizing seed production costs to be formulated and tested. This effort is significant because it is important to minimize the cost of seed production.
Approach/Background: Factors that affect growth and fermentation of NREL’s xylose-fermenting recombinant Zymomonas mobilis are being assessed to provide information for optimizing seed production for NREL’s current simultaneous saccharification and cofermentation process design. As such, synthetic media are being formulated with research-grade chemicals to mimic the composition of hardwood sawdust dilute acid pretreatment hydrolyzates. Work is being guided by the ethanol process development team to ensure that the research remains well aligned with NREL’s in-house process development objectives.
Status/Accomplishments: Work completed in FY 1996 showed that acetic acid is a potent inhibitor of seed production at the levels it is encountered in unconditioned (non-detoxified) hardwood hemicellulose dilute acid hydrolyzates. Effective seed production requires an acetic acid concentration below 7.5 g/L and is improved at lower acetic acid levels. Results also showed that inhibition by acetic acid can be minimized by controlling the fermentation pH at or near pH 6.0 and by supplementing with glucose. Work carried out during FY 1997 extended these findings to a different strain of xylose-fermenting Zymomonas than had been used in the previous work. Continuous culture work was also initiated in FY 1997 and will continue in FY 1998. Key accomplishments during FY 1997 included independently reproducing key steady-state, continuous-culture performance results previously obtained at NREL, as well developing improved procedures for starting up continuous cofermentations.
Publications and Presentations:
Summary Date: August 1997
Objective: To provide needed preventive maintenance and support to ongoing improvement projects in the 1 dry ton/day biomass-to-ethanol process development unit (PDU).
Approach/Background: The operation of the PDU is vital to successful commercialization of this technology. It verifies process performance, and generates performance and scale-up data necessary to design large-scale commercial facilities. This task is designed to maintain and upgrade the operability of the PDU and support and maintain utilities systems in the PDU that support other research efforts in the Alternative Fuels User Facility. Routine maintenance and calibration activities must be performed throughout the year. This maintains the operational readiness of the PDU for use by internal and external customers. In addition, this task plans and conducts activities that improve the operability of the pilot plant and enhances its capabilities to supply necessary process performance data for customers.
Status/Accomplishments: The following work is complete or is planned for completion by the end of the fiscal year. A database (which uses Maximo software) is being implemented to track PDU maintenance requirements. This ensures maintenance is performed in a timely manner, which is necessary to maintain operability of the plant. The process and instrumentation diagrams are being updated to reflect the current as-installed plant. The data acquisition and control system is being converted to Arcnet (a communications protocol) from the current serial communication. This is necessary to maintain compatibility with the control software and allows faster communication between the control instruments and the computers. A calibration program has been implemented to ensure periodic calibration of PDU instruments, which is necessary to ensure that accurate and high-quality data are generated by the PDU. A system has been ordered for handling and disposal of recombinant microorganisms. This system is necessary for external customers and for internal NREL work.
Summary Date: August 1997
Objective: Use an outside analytical testing laboratory to provide precise and accurate compositional analysis of routine lignocellulosic samples of interest to the Biofuels Program. This analytical information will supplement the more complex and/or non-routine analyses conducted in-house. The data will be used by Biofuels Program research groups to meet specific technical objectives defined in the annual operating plan.
Approach/Background: Routine feedstock, pretreated biomass, and the solid fraction of fermentation residues are to be analyzed for total solids, acid-insoluble and acid-soluble lignin, cellulose (as glucose), hemicellulosic sugars, starch, O-acyl groups, and ash. Pretreatment liquors and the liquid fraction of fermentation samples are to be analyzed for total and total dissolved solids, cellobiose, monomeric and total sugars, organic acids, glycerol, HMF, and furfural. During the course of these analyses, established Laboratory Analytical Procedures and the quality control protocols described in the Ethanol Project Quality Assurance Program must be followed. The results of each group of analyses are reported to the program for evaluation and data reduction.
Status/Accomplishment: A substantial number of samples were analyzed during FY 1996 and FY 1997. A total of 338 samples were analyzed during FY 1996 and an additional 163 samples, to date, have been analyzed during FY 1997. Assay reproducibility and analysis turnaround time consistently met or exceeded expectations. Improvements made to methods and reporting requirements have streamlined the process and have reduced the overall analysis cost per sample.
Summary Date: August 1997
Objective: To evaluate the preliminary technical and economic feasibilities of dilute-acid pretreatment methods for converting softwood forest thinnings to ethanol, and to identify major technical hurdles.
Approach/Background: Selective thinning of forests in the western United States will generate a large, sustainable quantity of softwood residues that may make an attractive feedstock for producing fuel ethanol. Softwood forest thinnings from Colorado and Northern California were selected as feedstocks for this project.
In contrast to hardwoods and herbaceous materials, published data on softwood conversion to ethanol are rather limited. The high lignin content coupled with the strong lignin-carbohydrate matrix make softwoods more difficult to pretreat. The major softwood species available in forest thinning operations in Colorado and Northern California (particularly in the Quincy area) are Douglas Fir, White Fir, and Ponderosa Pine. Based on a literature survey of softwood pretreatment methods for subsequent enzymatic conversion of the cellulose fraction to fermentable sugar, we focused on dilute sulfuric acid pretreatment as a promising option. Initially, preliminary pretreatment conditions were established for Douglas Fir, a predominant species available in the Colorado Front Range Forests. By mid-year, in support of a feasibility study for the Quincy Library Group (QLG) of near-term conversion technologies, work on the enzymatic conversion option was deferred in favor of a non-enzymatic, two-stage dilute acid hydrolysis process that uses a mixed feedstock of whole tree chips that consist of 70% White Fir and 30% Ponderosa Pine from Northern California. The ratio of the wood species selected is similar to that in forest thinning operations in and around the QLG area. Our view is that economical enzyme production technology is not yet available to meet the target of 2to 3-year time frame for a demonstration plant to produce ethanol from softwood forest thinnings.
Status/Accomplishments: Preliminary data indicate that these feedstocks can be readily converted to ethanol with dilute-acid pretreatment. However, additional developmental work is required to improve the ethanol yield, to identify equipment for unit operations, and to improve overall process economics.
Pretreatment and enzymatic hydrolysis of Douglas Fir: A series of experiments was carried out to establish preliminary pretreatment conditions for Douglas Fir chips. A typical pretreatment at 0.35% sulfuric acid concentration, 212oC, and 100 s solubilized 95% of the hemicellulose fraction and 22% of the cellulose. Overall sugar yields after enzymatic hydrolysis were 89% (cellulose) and 60% (hemicellulose). Subsequent treatment of the water-washed, acid-pretreated wood with dilute alkali at 60oC did not improve cellulose digestibility.
Two-stage dilute acid hydrolysis of mixed feedstock of whole tree chips: The mixed feedstock of whole tree chips, which comprised about 70% White Fir and 30% Ponderosa Pine, was used to establish processing parameters of a two-stage dilute sulfuric acid hydrolysis process. The sugar recovery yields are being determined.
Yeast adaptation to softwood prehydrolysates: The acid prehydrolysates of softwoods contain high concentrations of degradation products of hemicellulose and extractives. These compounds inhibit ethanol-fermenting yeast unless the hydrolysates are diluted with water to about 12% total solids from a concentration of about 25% solids that come from the pretreatment reactor. At 12% solids concentration, the ethanol yield obtained was 85% of theoretical. The dilution would result in higher energy requirement for ethanol recovery. Our target is to achieve 85% ethanol efficiency at 20% solids concentration. We have been able to adapt a Saccharomyces cerevisiae yeast strain to QLG softwood prehydrolysates at 17% total solids concentration. Also, we substituted expensive nutrients such as yeast extract and peptone with lower-cost ammonium phosphate.
Process economic evaluation: We developed a technoeconomic model of a two-stage dilute acid hydrolysis to identify key process areas that need to be improved to reduce the ethanol production cost.
Summary Date: September 1997
Objective: Develop advanced hemicellulose and cellulose hydrolysis methods for reducing overall costs of sugar release from lignocellulosic feedstocks for use in bioethanol production processes.
Approach/Background: Dilute-acid pretreatment processes have traditionally been directed at hydrolyzing the hemicellulose fraction of lignocellulosic feedstocks. The cellulose fraction is made more digestible by the pretreatment process and can be digested by cellulose enzymes to release fermentable sugars. The near-term costs of cellulase enzymes are prohibitive for use in commercial bioethanol production processes. Thus, thermochemical approaches for hydrolyzing cellulose are being investigated. The traditional concentrated acid approach to cellulose hydrolysis can achieve high yields of sugars, but suffers from high acid use and the need to cost-effectively recover and recycle the acid catalyst.
Dilute-acid approaches to achieving at least a partial level of cellulose hydrolysis have been developed as an extension of the countercurrent prehydrolysis processes that has recently been developed. Kinetic modeling of the hydrolysis reactions indicates that a countercurrent contacting of the solid biomass particles with the dilute-acid solution results in higher sugar yields with fewer sugar degradation products. Significant amounts of lignin are also solubilized in this approach, and render any remaining cellulose to be easily digestible by cellulase enzymes at relatively low enzyme loadings. The solubilization of lignin results in an increased level of toxicity of the hydrolyzate solution to ethanol-producing microorganisms, but detoxification approaches that remove the soluble lignin in a recoverable manner appear to offer unique opportunities for producing high-value coproducts for a variety of fuel additive and other applications.
Significant mass transfer and materials handling issues are likely hurdles in developing actual commercial-scale pretreatment processes based on the concept of countercurrent contacting of biomass particles with the dilute-acid catalyst solution. Thus, efforts are also under way to develop and test engineering-scale reactor systems that will closely resemble commercial-scale systems to demonstrate the feasibility of this approach.
Status/Accomplishments: A bench-scale continual shrinking bed reactor system that uses spring-loaded plungers to continually reduce the volume occupied by the remaining biomass as it is hydrolyzed has been developed and successfully tested with yellow poplar sawdust feedstock. The shrinking bed approach is necessary to reduce the volume of dilute-acid solution necessary to achieve significant levels of cellulose hydrolysis. Using 0.07% (w/v) dilute sulfuric acid in three temperature stages that range from 175· to 225oC, yields of 85% of glucose (monomers) from cellulose and 97% of xylose (monomers + oligomers) from hemicellulose at an overall sugar concentration of 3.6% (w/v) have been achieved. No cellulose or hemicellulose remains in the residual solids. Abut 70% of the lignin is also solubilized under these conditions.
A subsequent secondary heat treatment step was developed for near-quantitative conversion of xylose oligomers to monomers. A novel adsorption approach was then developed to remove soluble lignin in a form that can be used for boiler fuel or precursors for higher-value coproducts. The detoxified sugar solution can then be cofermented rapidly (<2 days) at high ethanol yields.
An engineering-scale (200-kg dry feedstock/day) prototype countercurrent reactor was designed and is being tested to determine whether this particular reactor design can achieve the necessary countercurrent contacting of solids and liquids in a manner that meets the low liquid volume requirements. Successful testing of this system will lead to the design and acquisition of an engineering-scale system that will be operated under actual reaction conditions.
Summary Date: September 1997
Objective: Develop pH-neutral hot-water pretreatment approach for hardwood sawdust feedstock and produce preliminary process flow diagram and mass balance.
Approach/Background: High-temperature hot water pretreatment approaches have been investigated for several years as alternatives to the dilute-acid pretreatment process. By using no added acid catalyst, there is a less corrosive environment within the pretreatment reactor, which may allow for less expensive materials of construction. It the past, most liquid hot water or steam pretreatment approaches with no added catalyst produced hydrolyzates in the pH range of 3–4. This is significantly lower than neutral conditions due to the release of acetic acid and other organic acids as hemicellulose is hydrolyzed. The general approach is to add a small amount of base to the liquid stream to keep the hydrolyzate at more neutral conditions (pH 5–7). The hypothesis is that the more neutral conditions will minimize sugar degradation reactions, particularly those that result from decomposition of xylose sugars released during the hot water pretreatment process.
Status/Accomplishments: The pretreatment of yellow poplar sawdust, which uses pressurized liquid hot water at temperatures higher than 220oC has been investigated. A 2-L Parr reactor agitated pressure vessel was used to conduct pretreatment experiments. A pH control loop with base addition was used. Conditions were identified that resulted in the greatest level of hemicellulose hydrolysis and the highest level of enzymatic digestibility of the pretreated solids. These batch data were used to develop a continuous flow-through reactor system and to identify possible large-scale equipment designs, subjects of a 4-month contract extension that is now nearing completion. The current process being evaluated is two-stage. The first stage is a countercurrent extraction of acetic acid from the incoming feedstock, which uses a very dilute base at 240oC. This is followed by a cocurrent, pH-controlled liquid hot water pretreatment at 240oC. Complete material and energy balances are currently being generated, along with a preliminary economic analysis that includes cost quotes for specialized pieces of equipment.
Summary Date: September 1997
Objective: Shake down and operate three NREL pre treatment systems to provide standard pretreated materials for NREL in-house and subcontractor research efforts.
Approach/Background: NREL currently has three pretreatment reactor systems and their ancillary equipment on location at Hazen Research, Inc. They are the 100-L paddle reactor, the 170-L Jaygo reactor, and a 2-in. diameter by 12-in. length stainless-steel percolation reactor. The systems have been installed. The intent of the larger systems is to provide large quantities of prehdrolyzed solids and prehydrolyzate liquors generated under well-established, routine conditions. These materials are provided as needed to support the specific needs of NREL in-house and subcontracted research efforts. The stainless steel percolation columns are used to evaluate liquid hot water flow-through pretreatment approaches.
Status/Accomplishments: A significant number of large-scale pretreatment runs have been conducted in the paddle reactor. Initial runs were made to establish solids loading limitations and validate performance data compared to bench-scale batch pretreatment reactors. After these conditions were investigated, the operation of the paddle reactor was limited to a number of specific requests to provide pretreated materials to various in-house and subcontract researchers.
The-170-L Jaygo and heating/cooling systems have been installed and mechanically shaken down. How ever, the heating/cooling system provided by the manufacturer does not meet the specifications required to heat up and cool down reactor contents. This system is in the process of being redesigned by the manufacturer.
A large number of hot water percolation pretreatment runs were conducted by Hazen, with direct experimental design input by NREL researchers. Yields of sugars and mass balance calculations were conducted for these experiments. The results indicate that xylose sugars are significantly degraded in the pH range that results from hot water pretreatment. The degradation can be controlled by reducing the residence time of the liquid. This is accomplished by flowing the hot water through at a rapid rate. However, the large volumes of water required at the higher liquid throughputs leads to an economically unattractive process.
Because of space limitations and other constraints, the pretreatment reactors at Hazen have been removed and are currently in storage at NREL. As demands for operating these reactors occur and as activities at Hazen permit, the reactors can be reinstalled and operated under the terms of a new subcontract. The current subcontract has expired.
Summary Date: September 1997
Objective: Determine the applicability of the Bioengineering Resources, Inc. (BRI) proprietary solvent extraction system for the removal of sulfuric acid, acetic acid and sugar degradation products from dilute acid prehydrolyzates.
Approach/Background: NREL and BRI are collaborating in Phase 2 of an STTR project to determine the applicability of BRI’s solvent extraction process to dilute-acid prehydrolyzates. This process was originally designed as an acid recovery and recycle step for use in concentrated sulfuric and hydrochloric acid processes, but work in Phase 1 demonstrated the possible use of this approach as a detoxification step for dilute-acid prehydrolyzates. The recovery and recycle of sulfuric acid may not be necessary in a dilute acid process, but effective removal of acetic acid and sugar degradation products from dilute-acid prehydrolyzates is necessary to allow for efficient fermentation to produce ethanol. In Phase 2, work on solvent toxicity on fermenting microorganisms, potential acetic acid coproduct recovery, overall process flow diagrams, and preliminary process economic evaluations will be conducted jointly by NREL and BRI.
Status/Accomplishments: The BRI STTR/CRADA project completed Phase 1 in May 1996. BRI received a Phase 2 award from DOE in August 1996, and a Statement of Work and CRADA modification has been negotiated for Phase 2 activities. Phase 2 will run for 24 months. The NREL Biofuels Program will participate in work funded both through BRI as a laboratory participant in the STTR program and in-kind work supported by Biofuels programmatic funds.
Phase 1 has involved investigating BRI’s proprietary solvent recovery system for sulfuric acid recovery in pretreatment processes, as well as preliminary evaluation of this system for removing acetic acid and sugar degradation products from dilute-acid prehydrolyzate. This represents an approach for detoxifying dilute acid prehydrolyzates. Results from the work on dilute-acid prehydrolyzate liquors indicated that the solvent systems effectively remove virtually all acetic acid, sulfuric acid, and many of the sugar degradation products from the liquor, which results in an extracted liquor of about pH 5. However, this liquor was not fermentable (with Saccharomyses cerevisiae yeast). Either lignin degradation products are assumed to still be present, or trace levels of solvents left behind still contribute toxicity to the fermentation system.
Phase 2 activities will include determining residual solvent toxicity on various fermenting microorganisms, optimizing the solvent system and extraction conditions to meet detoxification needs, potential recovery schemes that could allow for recovery of acetic acid as a potential coproduct, process engineering analyses to address acid recovery and detoxification applications, and preliminary evaluation of industrial interest for commercialization purposes. Preliminary fermentation of pure sugar solutions that had been contacted with solvent and processed through a solvent recovery system indicates that trace levels of residual solvent result in a significant time lag at the start of fermentations. Ultimate ethanol yield does not seem to be affected, but significantly longer fermentation times are required for the solvent-contacted sugar solutions.
Publications and Presentations: None.
Summary Date: September 1997
Objective: Understand the effects of operating variables on yields of xylose and glucose from hemicellulose and glucose from cellulose in NREL’s multi-stage dilute sulfuric acid hydrolysis process and provide analytical support to NREL in-house and subcontracted research staff to chemically analyze biomass samples.
Approach/Background: When lignocellulosic biomass is subjected to thermochemical treatment with dilute sulfuric acid, hemicellulosic sugars (primarily xylose) and glucose from cellulose are released. We can determine the relationship between particle size, temperature, residence time, acid concentration, physical shrinkage of the solid mass as a function of hydrolysis, and sugar yield by conducting experimental and modeling studies. This study uses yellow poplar and corn stover to establish kinetic models that predict yields of sugars through dilute sulfuric acid treatment. By incorporating the obtained kinetic models in the appropriate reactor design equations for a multi-stage, simulated countercurrent percolation process, the optimum operating conditions that give maximum yields of sugars at concentrations meaningful from a process economic standpoint, can be determined. Additionally, intraparticle and interparticle mass and heat transfer phenomena are being investigated.
Status/Accomplishments: We developed kinetic models that describe the release of xylose, oligomeric xylose, and glucose during dilute-acid treatment as a function of acid concentration, temperature, solid and liquid residence times, and the physical solubilization of the mass as a function of time. These models, based on experimental data and theoretical kinetic analysis, demonstrate that the use of a multi-stage, shrinking bed reactor scheme allows for variable temperature profiles to address the several kinetic carbohydrate species that greatly enhance the process performance. The models allow for optimization independently of yields and concentrations of all sugars.
Work is ongoing to mathematically optimize the various stages as to the types of reactor designs that can be used at each stage to yield favorable process economics. One goal of this modeling is to address the complexity of operation for near-ideal countercurrent movement of the solid and liquids as a function of hydrolysis by experimentally examining three reactor designs for the total hydrolysis process. These independently address (1) the easy hydrolyzable xylan; (2) the difficult fraction of xylan; and (3) the cellulose fraction. Once the three processes are optimized independently, the reactor sequence will be modeled sequentially.
Process economic modeling has been initiated that will use the kinetic results to guide further research efforts into reactor designs and operation for the total dilute-acid hydrolysis of lignocellulosic biomass.
Analytical service to NREL in-house and subcontracted research personnel was also provided. Liquid samples were analyzed for glucose, xylose, arabinose, galactose, mannose, acetic acid, furfural, hydroxyl-methyl-furfural, and solubilized lignin. Solid samples were analyzed for glucan, xylan, arabinan, galactan, mannan, Klason lignin, acid-soluble lignin, and total ash. The enzymatic digestibilities of certain samples were determined.
Summary Date: September 1997
Return to Table of Contents
File Created: October
13, 1997; Last updated: Thursday, 11-Nov-1999 10:23:30 EST