Lignin Derived Coproducts

Coproducts Development from Soluble Lignin Recovery Process

Research Funded by: U.S. Department of Energy Office of Fuels Development through the National Renewable Energy Laboratory

Project Manager: Robert Wooley 303.384.6825, Robert_Wooley@nrel.gov

Performing Organization: National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden, CO 80401-3393, http://www.nrel.gov

Principal Investigator:Bonnie Hames, 303.384.6345, bonnie_hames@nrel.gov

Contract Number: N/A

Contract Period: 10/97–09/98

Contract Funding:

FY 1998: $317,000

FY 1999: $0

Objective: A unique aspect of the countercurrent pretreatment approach is that a significant fraction of the lignin in yellow poplar sawdust is solubilized under the proposed reaction conditions. Soluble lignin is a component of the liquid hydrolyzate stream, as are sugars solubilized during the reaction. Although solubilized lignin in the hydrolyzates presents a challenge in terms of energy recovery to the process and potential toxicity to fermentative microorganisms, it also presents an opportunity as a potentially valuable intermed-iate in the production of chemical coproducts for fuel or nonfuel applications.

Approach/Background: Up to 70% of the lignin initially present in yellow poplar sawdust is solubilized in the countercurrent pretreatment. At a minimum, soluble lignin should be recovered in such a way that its heat content can be used as a boiler fuel, unless it can be converted to methane in the waste-water treatment step. Because this conversion has not been verified, work has continued to improve the performance and economics of the approach that adsorbs the soluble lignin upon a solid catalyst. Bench-scale data indicates that up to 90% of the solublized lignin can be removed by this approach.

Status/Accomplishments: The most effective approach for recovering the various lignin fractions that have been solubilized is to cool the hydrolyzate and filter out the soluble lignin that precipitates upon cooling (about one-third of the total soluble lignin). The remaining lignin can then be removed using the adsorbent. The adsorbed lignin can be removed by treating the adsorbent in a furnace, which allows recovery of the heat content of the solubilized lignin and regenerates the catalyst for reuse. In addition, fairly simple extractive methods can be used to remove the adsorbed lignin compounds in a manner such that the compounds can eventually be upgraded to fuel components. Initial work has indicated that the soluble lignin may be suitable for use in reformulated fuel components or oxygenates without the costly base-catalyzed depolymerization step that is needed to upgrade the insoluble lignin from a simultaneous saccharification and fermentation-based process.

The cost, performance, and lifetime of this adsorbent had previously been identified in process economic evaluations as a key cost component of the process. Thus, strategies that improve the performance and lifetime, and that identify possible uses of spent adsorbent that would provide a salvage value, have been investigated. We identified adsorbent formulations that improve the adsorption rate and lignin removal efficiency. Regeneration experiments suggest that regeneration cycles can be increased from 100 to 500 or more. Industrial suppliers have suggested that agreements can be made for a regeneration credit of 30% of the original price of the adsorbent. The demon-strated improvements have reduced the cost of the soluble lignin adsorption step from $0.18/gal ethanol to $0.01/gal ethanol (including salvage value credit).

In addition to upgrading coproducts to fuel use, the various lignin fractions from the countercurrent pretreatment process have other potential nonfuel applications. Examples of potential coproducts include metal chelators, compounds similar to specialty lignosulfonates, and industrial chemicals such as phenol, catechol, and vanillin. Although these products may not have as large a market as fuel-use coproducts do, their higher value may compensate. The nonfuel uses may provide an opportunity to demonstrate these process concepts at a commercial scale and to ultimately enable the commercialization of a bioethanol process based upon this reaction concept.

Publications and Presentations: None

Summary Date: March 2000

Upgrading Nonfermentable Compounds Solubilized During Dilute- Acid Pretreatment of Lignocellulosics to High Octane Fuel Additives

Research Funded by: U.S. Department of Energy Office of Fuels Development through the National Renewable Energy Laboratory

Project Manager: Robert Wooley 303.384.6825, Robert_Wooley@nrel.gov

Performing Organization: Department of Chemical Engineering University of Sherbrooke, Sherbrooke, Quebec J1K 2R1, Canada

Principal Investigator: N. Abatzoglou, 819.821.7904, Nicolas@gcm.usherb.ca

Contract Number: ACI-9-29051-01

Contract Period: 06/99–06/00

Contract Funding:

FY 1999: $66,858

Objective: The objective of this project is to convert the low-molecular-weight lignins and pentoses found in dilute-acid pretreatment liquors to high-octane hydrocarbons and oxygenated fuels other than ethanol and methanol.

Approach/Background: NREL contracted with the University of Sherbrooke to perform two distinct tasks related to the components found in dilute-acid pretreatment liquors:

1. to study the upgrading of low-molecular-weight lignins; and
2. to review existing technologies for upgrading pentoses.

The desired products are high-octane hydrocarbons or oxygenated fuels or fuel additives other than ethanol or methanol.

The goal of Task 1 is to selectively cleave C-O-C ether bonds between constitutive oxyaromatic moieties present in a ‘low- molecular-weight lignin substrate,’ while pre-serving the aromatic ring, to produce green fuel additives. Thermal hydrocracking, proceeding through free radical mechanisms at high temperatures and pressures (endothermic formation of radicals + exothermic hydrog-enation) is not selective. It destroys a large proportion of aomatic rings and causes coke-forming polymerization and condensation reactions. Consequently, catalytic hydrocracking has been chosen as the route to the desired products. The challenge is to find a catalytic system that better selects for cracking and hydrogenation of the ether bonds rather than the aromatic ring. Mo pseudohomogeneously dispersed in a solid catalyst has been chosen for the selective cleavage of C-O-C bonds followed by hydrogen addition. The catalytic hydrocracking mechanism is ionic and is based on the chemistry of the carbonium cation.

Currently the pentoses solubilized during pretreatment of biomass are planned to be cofermented with glucose to make ethanol, using Zymomonas mobilis or other organisms. An alternative strategy would be to convert the pentoses to furfural and then selectively hydrogenate them to furans, which have a very high octane number (about 200) and relatively low boiling points (31^o–94^oC). A substantial amount of work has already been performed on the conversion of pentoses to furfural and on the hydrogenation reaction. A literature review will assess the problems of this approach and whether it is worthwhile to reinvestigate this strategy.

Status/Accomplishments: A final decision on the most appropriate catalytic system to complete Task 1 is imminent. Reaction conditions for achieving considerable lignin conversion to useful products will be identified by the end of the present contract.

During this first year of research and development, the following variables and parameters have been the focus of work:

• Composition of the Mo catalyst
• Nature and use of solvents of and liquid carriers for the lignin substrate
• Sources and preparation of the catalyst
• Ways of distributing the catalyst in the reactive phase
• Use of a batch or semibatch reactor for the hydrocracking reaction (flowing or static H₂ blanket)
• Reaction temperature
• Reaction time
• Hydrogen pressure

The results obtained so far indicate the following:

• The low molecular weight lignin substrate is more responsive to the pseudohomogeneous catalytic hydrocracking than the original lignin.
• Qualitative experiments using anisole showed that the Mo organometallic (Mo(CO)₆) has catalytic activity in terms of hydrocracking.
• Mo(CO)₆, was less active than Ni and Co organometallics in terms of hydrocracking. However, it proved quite effective, especially when activated with in situ nascent sulfur. Conversions to ether extractables, which represent the interesting products, are around 50% of the substrate.
• Reaction times as short as 30 min have proved effective. Lower residence times are currently under investigation.
• Carrier liquid phases have proved problem-atic because of interference with the reaction chemistry.
• Pressure levels are maintained between 600 and 1000 psia. No definite conclusions have been drawn regarding the ideal reaction pressure. Our conclusions so far are that these pressure levels are appropriate and that product yields will not increase at higher pressures. We believe that hydrogenation is not the rate-controlling step. Rather, we believe that diffusion phenomena plus carbonium formation steps are the steps to optimize.

Publications and Presentations: None

Summary Date: January 2000

Lignin Conversion to Fuels

Research Funded by: U.S. Department of Energy Office of Fuels Development through the National Renewable Energy Laboratory

Project Manager: Robert Wooley 303.384.6825, Robert_Wooley@nrel.gov

Performing Organization: National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden, CO 80401-3393, http://www.nrel.gov

Principal Investigator:E. Chornet, 303.384.6240, esteban_chornet@nrel.gov

Contract Number: DE-AC02-83CH10093

Contract Period: 10/1/97–9/30/99

Contract Funding:

FY 1998: $342,614

FY 1999: $748,697

Objective: Develop processes to convert lignins into hydrocarbon and oxygenated high-octane fuel additives compatible with reformulated gasoline (RFG).

Approach/Background: Lignin is present (15–30 wt%) in all lignocellulosic biomass. Any ethanol production process will have lignin as a residue. At present it is envisaged that lignin will be combusted to provide heat and/or power for the ethanol process. Increasing the value of this residue can significantly enhance the competi-tiveness of biomass-to-ethanol conversion.

Given the molecular nature of lignin it is conceivable to convert lignin into added value end products, especially fuel additives. Current specifications for gasoline define the properties of the products we are targeting. Gasoline can contain only low levels of benzene (<1%), a controlled amount of aromatics (<25%), and olefins (<5%), but should contain a low level of oxygen (about 2%). The gasoline boiling point (90% should distill below 149^oC) and Reid vapor pressure (should be 48–56 kPa) are also important parameters. Octane number deter-mines gasoline fuel grade and ranges from 85 to 93. Valuable additives should therefore be hydrocarbons or oxygenates that have a high octane number (>100), and are compatible with the gasoline boiling point and Reid vapor pressure parameters. This project is thus targeted to obtain the necessary scientific and techno-logical data with which to evaluate the economics and markets for converting lignin into valuable fuel additives.

The first step to converting lignin into a product compatible with the transportation fuel market is to decrease its molecular weight. Currently, base-catalyzed depolymerization of lignin is the process being studied to achieve this. Professor Shabtai's group at the University of Utah (University of Utah) have utilized bases (KOH, NaOH, Ca(OH)₂, etc.) in alcoholic (e.g., methanol, ethanol) or aqueous solution, and relatively high temperatures (250^o–40^oC) to accomplish lignin depolymerization. Techno-economic analyses have indicated that the use of alcohols is prohibitively expensive due to incorporation of the alcohols in the product, parasitic reactions that consume the alcohols, and the energy involved in recycling the alcohols. Currently work on the depolymer-ization reaction is focused on using aqueous solutions with low concentrations of NaOH or recyclable solid bases.

Further processing is necessary before the product of lignin depolymerization is suitable for fuel applications. The product must be either partially or completely deoxygenated depending on whether an oxygenate or hydrocarbon product is the final target. In one approach, the base-catalyzed depolymerization step is followed by upgrading via hydro-processing, involving catalytic hydrodeoxygen-ation, and then hydrocracking. The base-catalyzed depolymerization product is thus con-verted into a mixture of aromatic hydrocarbons. In a second approach, selective hydrotreating is applied in which oxygen-containing functional groups are preserved while C-C bonds are cleaved to increase the yield of monomeric phenols. The hydrotreating step is followed by etherification to give aromatic ethers whose performance should be at least as good as current commercial oxygenated fuel additives.

A team involving personnel from the University of Utah, Sandia National Laboratories (SNL), and NREL is working to further the development of a process for making fuel additives from lignin. The University of Utah group is studying the fundamentals of lignin depolymerization and hydrotreating. SNL is studying lignin depolymerization with solid alkaline earth oxides. The role of the NREL group is to supply lignin feedstocks and to chemically characterize the feedstocks and products generated by the other two groups. Dr. Esteban Chornet at NREL is coordinating the work of the team.

Status/Accomplishments: Systematic studies have been performed at the University of Utah on the base-catalyzed depolymerization step using stirred autoclaves. The effects of reaction time, temperature, and pressure, base type, and concentration, and the solvent used, have been examined. High yields of depolymerized lignin (60%–80%) are obtained at moderate temperatures (320^o–330^oC) and short reaction times (5–15 min) in aqueous media. At the University of Utah, most of the work has been performed with NaOH, however, while KOH is equally effective Ca(OH)₂ is not. Solid bases, such as a Cs exchanged X-type zeolite, have also been found to be effective. The deploy-merization reaction is currently being scaled up in a flow reactor capable of processing 100 g of lignin per hour.

Depolymerized lignin samples have been hydroprocessed at the University of Utah into mixtures of aromatic and naphthenic hydrocarbons that could be used as fuel additives. Analyses of these hydrocarbon samples, performed at NREL, indicate that they are predominately (~65%) made up of aromatic hydrocarbons and that so far only 25% of the product is in the gasoline boiling range (BPt. < 200^oC). Estimates indicate that the octane of these products is in the 100–110 range. Production of hydrocarbon products is being scaled up so that sufficient is available for fuel- property testing.

SNL has been studying the kinetic and reaction chemistry of lignin depolymerization and have been developing a CaO/NaOH catalyst system utilizing rapidly heated batch microreactors. SNL researchers have replicated results from the University of Utah showing that high yields of depolymerized lignin can be obtained from an aqueous NaOH system. In addition they have found that combinations of NaOH with CaO or MgO show positive synergies giving higher conversions in combination than when used alone. The data suggests that the synergy results from the combined system acting to limit neutralization of the base catalyst. A sample of depolymerized lignin made with the CaO/NaOH system was hydroprocessed under the conditions normally used at the University of Utah and found to give a product that was very similar to that produced by the University.

A process flow sheet has been constructed for the process of converting lignin into a high-octane hydrocarbon fuel additive. From this the production cost for the fuel additive has been estimated at about $1.06/gal. This estimate assumed the plant received a fraction of the lignin (600 tpd) from an ethanol plant. Performance targets used in the calculations included that 100% of the lignin in the SSF residue could be solubilized in a relatively low concentration aqueous NaOH stream with an 8% lignin concentration.

The cost of base recovery should be minimized by limiting the NaOH concentration to 1% or less in the lignin depolymerization step. The overall yield of hydrocarbon product in the gasoline boiling range was assumed to be 70% of the theoretical yield, equivalent to about 50% on a mass basis. Extraction of the depolymerized lignin intermediate should be performed with a relatively low cost solvent such as toluene. Current research is aimed at meeting these performance targets.

The value for a high-octane (R + M/2 = 110) hydrocarbon fuel additive has been estimated based on the Annual Energy Outlook projection of $20–$25/bbl for crude oil in 2010, and that the value of an octane boosting additive is in the range of $00.7–$01.4 per octane gallon. The value of hydrocarbon fuel additive should be in the range of $0.97–$01.14 per gallon based on these assumptions. Thus if performance targets can be met or slightly exceeded it should be possible to utilize the lignin from a bioethanol plant to enhance the competitiveness of biomass-to-ethanol conversion.

Publications and Presentations:

1. Shabtai J., W. Zmierczak, S. Kadangode, E. Chornet, and D.K. Johnson. 1999. Lignin Conversion to High-Octane Fuel Additives. p. 811-818. Proceedings, Fourth Biomass Con-ference of the Americas: Biomass—A Growth Opportunity in Green Energy and Value-Added Products. Edited by R.P. Overend and E. Chornet. Pergamon/Elsevier Science, Oxford, U.K.
2. Shabtai J. S., W. Zmierczak, E. Chornet, and D. K. Johnson. 1999. Conversion of Lignin. 2. Production of High-Octane Fuel Additives. ACS Div. Fuel Chem. Preprints, 44:267–272.

Summary Date: January 2000