JUST BECAUSE OIL WELLS will run dry one day does not mean the world is fated to run out of gasoline. A new biomass-based industry supported by venture capitalists, government funding, and big oil is rising to convert plant sugars (carbohydrates) into conventional liquid transportation fuels (hydrocarbons) that can be used to boost fuel supplies until a permanent renewable transportation solution can take hold—namely solar- or hydrogen-powered electric cars.
Several technology breakthroughs announced during the past year are signaling that the chemical industry is in for a period of rapid growth in biogasoline and organic compounds that can substitute for gasoline. The reason for these developments, scientists in the biofuels arena are saying, is that the energy, environmental, and economic benefits and impacts of first-generation biofuels—ethanol from corn and biodiesel from vegetable oil or animal fat—leave room for improvement.
"Crude oil has been used for more than 100 years, and oil companies have optimized processes for converting petroleum to fuels," says George W. Huber, an assistant chemical engineering professor at the University of Massachusetts, Amherst. "We now need to do the same for biomass." Huber has joined a growing cadre of scientists who are building or redefining their careers by conducting biofuels research.
"In the future, consumers likely won't know if they are putting biofuels in their tanks or not, because the chemical composition of the fuels is essentially the same as petroleum-derived fuels," Huber notes. "The challenge for chemists and chemical engineers is to efficiently produce liquid fuels from biomass while fitting the technology into the existing refinery and transportation fuels infrastructure."
Converting soluble plant-derived sugars or raw cellulosic biomass from dedicated crops, agricultural residues, or municipal waste into liquid fuels requires removing most or all of the oxygen atoms in the feedstock—cutting the "hydrate" out of carbohydrate as carbon dioxide and water to form products that are more amenable to combustion. This deoxygenation must be accompanied by standard refinery-type reactions to "upgrade" the intermediate products into branched hydrocarbons and aromatic compounds for gasoline and into longer chain hydrocarbons for jet and diesel fuels.
One company that has a jump on making these biofuels is Virent Energy Systems, in Madison, Wis. The company was founded in 2002 by University of Wisconsin chemical engineering professor James A. Dumesic and research scientist Randy D. Cortright to commercialize an "aqueous-phase reforming" process to produce hydrogen from glucose and other water-soluble, plant-derived sugars (C&EN, Sept. 2, 2002, page 29).
Interest in hydrogen as a fuel has temporarily waned because of lingering technology challenges, but Virent has been working to modify aqueous-phase reforming and combine it with refinery reactions. The outcome is an integrated process, which Virent calls BioForming, to produce gasoline, jet fuel, diesel fuel, and other chemicals.
In September, Virent announced that it had been granted several patents for this sugar-to-liquid fuels strategy. Virent had already licensed Dumesic's original aqueous-phase reforming patents from the Wisconsin Alumni Research Foundation, a nonprofit technology-transfer organization that administers University of Wisconsin inventions.
BIOFORMING UTILIZES proprietary heterogeneous catalysts at moderate temperatures (80−400 °C) and pressures (up to 50 atm) in a collection of parallel and tandem reactions, explains Cortright, who is now Virent's chief technology officer. In one process the soluble carbohydrate feedstock is first pretreated by hydrogenation over a ruthenium catalyst supported on carbon (Ru/C) to form intermediate sugar alcohols (polyalcohols). Part of the feedstock separately undergoes hydrogenolysis (hydrogen-terminated C???C bond cleavage) to form shorter chain oxygenated compounds such as glycerol and propylene glycol.
After pretreatment, the streams pass through a reactor for the aqueous-phase reforming step, which is a mix of chemical reactions rolled into one, Cortright says. The compounds react with water over a platinum-rhenium catalyst supported on carbon (Pt-Re/C) to form H2, CO2, C1–C4 alkanes (methane to butane), and a collection of C1–C6 alcohols, ketones, aldehydes, organic acids, and cyclic compounds. The H2 is used in situ during the reaction to create the C1–C6 monooxygenated products, he notes, with excess H2 shunted off to other BioForming steps.
One limitation of aqueous-phase reforming, and of converting carbohydrates to fuels in general, is that a significant portion of the carbon is lost as CO2, compromising the yield of hydrocarbons. Biomass-to-fuels technologies are in theory nearly "carbon neutral" if renewable biomass is used as the sugar source because a nearly equivalent amount of CO2 is taken up by the plants as they grow relative to the amount of CO2 emitted when the fuel is burned. For Virent and other biofuel producers, their processes could help reduce global atmospheric CO2 levels if some or even all of the by-product CO2 is trapped and then used as a feedstock in other processes or is sequestered.
As for the other products of BioForming, the C1–C4 alkanes are burned to generate heat to run the reactors, and the C1–C6 monooxygenated compounds are upgraded to oxygen-free hydrocarbons in a continuous process by using catalytic refinery condensation and hydrotreating techniques with tungsten and zeolite catalysts, Cortright says.
For gasoline, a blend containing primarily C5–C10 alkanes and aromatics is made by using the aluminosilicate zeolite ZSM-5. Virent has demonstrated a complete pilot-scale sugar-to-gasoline process in a reactor system containing four different catalyst beds (Ru/C, Pt-Re/C, W/ZrO2, and ZSM-5) operating at the same pressure and with no intermediate separations required. Currently, the product distribution based on carbon content is about 25% CO2, 25% C1–C4 alkanes, and 50% gasoline. About 60% of the H2 needed in the various hydrogenation reactions is generated during the aqueous-phase reforming step.
"Virent has proven that sugars can be converted into the same hydrocarbon mixtures of today's gasoline blends," Cortright says. "Our products match petroleum gasoline in functionality and performance."
Meanwhile, Dumesic's group at Wisconsin has continued to work independently of Virent to devise chemistry to upgrade aqueous-phase reforming products. Dumesic and coworkers have published several papers on the research, which culminated in a report published last month describing an integrated process using flow-through reactors to prepare gasoline, jet fuel, and diesel fuel blends (Science 2008, 322, 417).
The chemistry uses the same Pt-Re/C catalyst for aqueous-phase reforming as Virent's process. The subsequent refining steps are similar to those in Virent's BioForming process, with differences primarily being in the choice of catalysts. Dumesic's ongoing developments will likely end up in the hands of Virent via the Wisconsin Alumni Research Foundation, he says.
CONVERTING SUGARS, cellulose, and eventually raw lignocellulosic biomass to hydrocarbons suitable for transportation fuels in a cost-effective manner "is a very tough challenge, as there are a number of chemical transformations that need to take place," comments Leo E. Manzer, president and chief executive officer of Wilmington, Del.-based consulting firm Catalytic Insights. "Virent and Dumesic have made significant progress in simplifying the overall process; it's a major step forward."
Manzer, who was a leading research scientist at DuPont before retiring in 2005, currently focuses on developing catalysts for biomass conversion to fuels and chemicals. "I fully expect that Virent and Dumesic will be able to consolidate the reaction steps and develop a multifunctional catalyst that might do all the chemistry in a single reactor—that is, sugars in and hydrocarbons out," Manzer says.
"Oil companies have optimized processes for converting petroleum to fuels. We now need to do the same for biomass."
Virent currently uses processed cane sugar for pilot-scale operations, Cortright says, but any type of water-soluble carbohydrate will do: sucrose (from cane or beets), glucose (from corn), fructose (from fruit), or sugar alcohols such as sorbitol and mannitol. That versatility contrasts with the fermentation process used to make ethanol, which is limited to working with one specific type of sugar at a time, Cortright points out. Eventually, BioForming will be able to start with raw biomass, he notes. It's an advantage that will allow Virent to utilize the lowest cost feedstock available in different regions, he says.
BioForming also avoids dependence on microbes used in fermentation processes, which can be finicky and time-consuming, Cortright adds. Batch fermentation to make ethanol takes at least two days, whereas in Virent's BioForming process the reaction time is about one hour, he says. In addition, BioForming needs a simple phase separation to isolate gasoline from the aqueous phase: "The fuel floats on water, so it's easily isolated," he notes. Ethanol requires an extensive, energy-intensive distillation to remove water.
Cortright says Virent's analyses based on a 100 million-gal-per-year commercial plant indicate that its production costs of biogasoline can be competitive with petroleum-based gasoline at crude oil prices above $60 per barrel. That works out to about $1.80 per gal for Virent biogasoline versus $2.12 per gal for petroleum-derived gasoline, and it's about 20 cents per gal cheaper than bioethanol production costs.
In March, Virent and Royal Dutch Shell announced a strategic alliance to move the BioForming process from the pilot stage to commercial implementation. Agribusiness giant Cargill and automaker Honda are also supporting Virent's efforts.
"Virent is moving rapidly to prove the technology," Cortright says. BioForming is a potential "game-changing technology" to help address global energy supply and environmental concerns, he emphasizes.
TAKING A DIFFERENT approach to green gasoline, scientists at Gevo, an Englewood, Colo., biofuels firm, use a genetically engineered microbe to produce 2-methylpropanol, which is known in the industry as isobutanol.
This compound is a chemical intermediate that can be used as a gasoline ingredient, "but the real potential of 2-methylpropanol is that it can be dehydrated to isobutylene and then converted into octane, aromatics, and other gasoline ingredients," says Patrick R. Gruber, Gevo's CEO. Gruber previously served as chief technology officer of NatureWorks, a joint venture of Cargill and Dow Chemical (now Cargill and Teijin) formed to make corn-derived polylactic acid polymers used for packaging and textile fibers.
Gevo was founded in 2005 by California Institute of Technology chemical engineer Frances H. Arnold, chemist Matthew W. Peters, and molecular biologist Peter Meinhold to reap the benefits of their directed-evolution protein engineering techniques for making modified microbes for biocatalysis. The company started out to develop fermentation technology for converting methane to methanol. Gevo later acquired an exclusive license to a method developed by chemical engineer James C. Liao and coworkers at the University of California, Los Angeles, for modifying Escherichia coli to produce butanols or other short-chain alcohols (C&EN, Jan. 7, page 21).
An existing route to butanol is the acetone-butanol-ethanol (ABE) industrial fermentation process, which uses the bacterium Clostridium acetobutylicum to make 1-butanol from sugars. It dates back to the early 20th century, before the more efficient petrochemical route to make 1-butanol from propylene was invented. But C. acetobutylicum is difficult to grow, not easy to genetically manipulate, and gives the ABE mix of products, not just butanol.
Liao's group turned to the biotechnology workhorse E. coli, which grows quickly and is easy to modify to produce a single product. Although E. coli doesn't naturally produce 1-butanol, Liao and coworkers reasoned that modifying the bacterium's amino acid biosynthetic pathway with snippets of C. acetobutylicum genes could produce enzymes that selectively convert sugars into 1-butanol, 2-methylpropanol, or other alcohols.
Acquiring the butanol technology created an instant opportunity for Gevo, Gruber says. The company is now working full tilt to commercialize 2-methylpropanol production and the additional chemistry needed for the conversion to gasoline. To that end, Gevo has developed a continuous production process dubbed Gevo's Integrated Fermentation Technology (GIFT).
GIFT overcomes all of the problems that plague ABE production, Gruber says. The bacterium produces only 2-methylpropanol at rates and yields similar to ethanol fermentation but at concentrations above the compound's solubility in water, which simplifies its isolation and purification by phase separation.
"We essentially have a four-carbon building block that we can use to make any type of fuel using traditional refinery reactions," Gruber says. He envisions using GIFT to retrofit existing ethanol plants and to build new 2-methylpropanol plants in strategic locations near biomass sources. "Done right, it will have very low capital cost, low operating cost, low technology risk, and many product opportunities," Gruber believes.
With GIFT, Gevo can retrofit existing ethanol plants to make 2-methylpropanol at a capital cost of about 20 to 30 cents per gal, Gruber points out. Adding the hydrocarbon upgrading units would cost an additional 20 to 30 cents per gal, making the total capital investment to make hydrocarbons just 40 to 60 cents per gal. With the retrofits, the production costs would be competitive with petroleum-based gasoline when oil is $50 per bbl or more, Gruber says.
Gevo has partnered with engineering and construction firm ICM, based in Colwich, Kan., to begin commercializing the technology. ICM has designed or built about two-thirds of the existing ethanol plants in the U.S., Gruber notes. Gevo is now using microbes in a 10,000 gal-per-year pilot plant to produce 2-methylpropanol, he says. By mid-2009, the company should be operating a 1 million-gal-per-year plant in Missouri made by retrofitting an ethanol facility. And by 2011, Gevo should have its first commercial-scale plant on-line that can produce 20 to 50 million gal of 2-methylpropanol and hydrocarbons per year.
Gevo is not the only company chasing after butanols. DuPont scientists previously developed genetically modified E. coli to make 1-butanol, and the company is extending the technology to include microbes that make 2-butanol and 2-methylpropanol.
In June 2006, DuPont and BP announced a partnership to commercialize the technology. The companies are building research labs and a 5,000 gal-per-year pilot-scale plant in the U.K. to produce the alcohols. The facility, which is expected to be operational in early 2010, is being built on the same site as a 110 million-gal-per-year ethanol plant that also is scheduled to go on-line in 2010. The ethanol facility might be converted to produce 1-butanol or 2-methylpropanol once that technology is ready, the companies have said. Both Gevo and DuPont have filed numerous patents covering their respective technologies, but neither company has been granted patents as yet.
On another front, the University of Massachusetts' Huber and graduate students Torren R. Carlson and Tushar P. Vispute reported earlier this year a continuous catalytic pyrolysis method that permits direct conversion of purified cellulose into aromatic compounds that can be used to make gasoline (C&EN, April 21, page 10). The group has since extended the process to start with wood chips, sugarcane bagasse, and corn stover. Huber is a former Dumesic graduate student at Wisconsin and worked on developing aqueous-phase reforming.
Pyrolysis is a process that uses medium heat and low oxygen conditions to partially break down cellulose, plant biomass, or any type of carbon-containing agricultural, animal, or industrial waste into "biocrude," an oil that can be used as a fuel oil (kerosene) or upgraded to gasoline-range hydrocarbons. The UMass researchers sorted out the necessary reaction conditions to control the pyrolysis of cellulosic material mixed with fine particles of the zeolite catalyst ZSM-5 to directly make the gasoline compounds in a single step.
The cellulose first decomposes into a mix of more than 300 liquid hydrocarbons, along with by-products that include coke, H2O, CO, and CO2. The biocrude subsequently enters the zeolite's pores and undergoes a series of decarbonylation, dehydration, oligomerization, and other reactions, Huber says. The process, which Huber has named catalytic fast pyrolysis, takes less than two minutes to complete at 600 ºC in a specialized reactor and generates a set of aromatic compounds, including benzene, toluene, xylenes, and naphthalene.
The process has one shortcoming in that it currently produces only aromatics. But Huber notes that the aromatics are actually more valuable than gasoline. As fuel ingredients they can be blended with alkanes to make gasoline, or they can be hydrogenated to produce alkanes. "If we combine a hydrogenation step with our pyrolysis process, then we can in principle make a complete gasoline," he notes.
The current challenge, however, "is to scale up the technology," Huber says. To accomplish that, a company named Anellotech has just been formed. He estimates that it might take five years for Anellotech to get a pyrolysis-derived biofuel to the pump.
THE INCREASED INTEREST in pyrolysis is sparking several industrial developments. In September, refinery technology provider UOP teamed up with Canada-based Ensyn to develop a process to pyrolyze cellulosic biomass to biocrude that can be refined into gasoline and other fuels. Ensyn already operates several pyrolysis facilities that use its Rapid Thermal Processing technology to convert forest and agricultural wastes to kerosene and related fuels for boilers and turbines. According to UOP's Jennifer Holmgren, director of the renewable energy and chemicals program, the venture should lead to commercial quantities of transportation fuels within three years.
In analyzing the liquid-phase processing and pyrolysis approaches, associate chemistry professor Mark Mascal of UC Davis sees the value of the chemistry to produce biogasoline and other fuels, but notes that "there is a price to pay."
Aqueous-phase reforming still has the same feedstock limitations that ethanol faces, he says, and the initial products must undergo condensation reactions to lengthen the hydrocarbon chains to make them suitable as fuels, which adds complexity and cost. Plus, the overall yield of hydrocarbons is low. The pyrolysis approach doesn't have the same feedstock constraints because it can already start from raw biomass, Mascal adds. But it requires further reaction chemistry to upgrade the biocrude, or in Huber's case, the aromatics, neither of which can be used directly as a transportation fuel.
Considering these limitations, he suspects aqueous-phase reforming and pyrolysis might not be the best solutions in the long run, "being ultimately surpassed by an even newer generation of cheaper biofuels that also are compatible with the current automotive infrastructure," Mascal says.
Among the more interesting prospects, he observes, are the butanols being developed by Gevo and the DuPont-BP venture; biodiesel produced by farming algae; and substituted furans (furfurals) being pursued by several companies and academic teams, including Dumesic's group and Mascal's own group. These compounds all have an energy content equal to or better than gasoline's and much higher than that of ethanol, Mascal points out.
Earlier this year, Mascal and postdoc Edward B. Nikitin reported a method to directly convert cellulose into 5-(ethoxymethyl)furfural (C&EN, Aug. 11, page 37). "The strength of our approach to furanic fuels," Mascal says, "is that it can operate directly on raw biomass with nearly quantitative yield and uses nothing more sophisticated than hydrochloric acid." As a diesel fuel ingredient, 5-(ethoxymethyl)furfural is already on the path to commercialization by Avantium, a Netherlands-based technology development company spun off by Shell.
Another one of these furanic compounds is γ-valerolactone, which is being championed by chemistry professor István T. Horváth of Eötvös University, in Budapest, Hungary (Green Chem. 2008, 10, 238). Horváth's group has created an integrated catalytic process to convert sucrose to γ-valerolactone and on into a mixture of fuel-range compounds (Top. Catal. 2008, 48, 49). He admits that furanic fuels likely will not fit directly into the current transportation infrastructure as well as biogasoline can, but Horváth believes that over time they could be the most economical biofuels because the production process moves from biomass to usable fuel with minimal chemical upgrading.
Sugar-based ethanol and biodiesel from vegetable oil or animal fats are "necessary starting points on the way to renewable fuels," Catalytic Insight's Manzer says. But eventually, a renewable hydrocarbon product obtained from lignocellulosics, the most abundant organic material in the world, that fits into the existing transportation infrastructure is needed.
"There does not have to be a single solution to this problem," he emphasizes. "With a U.S. gasoline market opportunity of 160 billion gal per year, there is plenty of room for many different cost-effective approaches—I don't see any one approach as being truly unique, and no single company can supply such a huge market. We are all looking for energy independence, so I am hoping all these large and small companies are successful."