On a blackboard, it looks so simple: Take a plant and extract the cellulose. Add some enzymes and convert the cellulose molecules into sugars. Ferment the sugar into alcohol. Then distill the alcohol into fuel. One, two, three, four — and we’re powering our cars with lawn cuttings, wood chips, and prairie grasses instead of Middle East oil.
Unfortunately, passing chemistry class doesn’t mean acing economics. Scientists have long known how to turn trees into ethanol, but doing it profitably is another matter. We can run our cars on lawn cuttings today; we just can’t do it at a price people are willing to pay.
The problem is cellulose. Found in plant cell walls, it’s the most abundant naturally occurring organic molecule on the planet, a potentially limitless source of energy. But it’s a tough molecule to break down. Bacteria and other microorganisms use specialized enzymes to do the job, scouring lawns, fields, and forest floors, hunting out cellulose and dining on it. Evolution has given other animals elegant ways to do the same: Cows, goats, and deer maintain a special stomach full of bugs to digest the molecule; termites harbor hundreds of unique microorganisms in their guts that help them process it. For scientists, though, figuring out how to convert cellulose into a usable form on a budget driven by gas-pump prices has been neither elegant nor easy. To tap that potential energy, they’re harnessing nature’s tools, tweaking them in the lab to make them work much faster than nature intended.
While researchers work to bring down the costs of alternative energy sources, in the past two years policymakers have finally reached consensus that it’s time to move past oil. The reasoning varies — reducing our dependence on unstable oil-producing regions, cutting greenhouse gases, avoiding ever-increasing prices — but it’s clear that the US needs to replace billions of gallons of gasoline with alternative fuels, and fast. Even oil industry veteran George W. Bush has declared that “America is addicted to oil” and set a target of replacing 20 percent of the nation’s annual gasoline consumption — 35 billion gallons — with renewable fuels by 2017.
But how? Hydrogen is too far-out, and it’s no easy task to power our cars with wind- or solar-generated electricity. The answer, then, is ethanol. Unfortunately, the ethanol we can make today — from corn kernels — is a mediocre fuel source. Corn ethanol is easier to produce than the cellulosic kind (convert the sugar to alcohol and you’re basically done), but it generates at best 30 percent more energy than is required to grow and process the corn — hardly worth the trouble. Plus, the crop’s fertilizer- intensive cultivation pollutes waterways, and increased demand drives up food costs (corn prices doubled last year). And anyway, the corn ethanol industry is projected to produce, at most, the equivalent of only 15 billion gallons of fuel by 2017. “We can’t make 35 billion gallons’ worth of gasoline out of ethanol from corn,” says Dartmouth engineering and biology professor Lee Lynd, “and we probably don’t want to.”
Cellulosic ethanol, in theory, is a much better bet. Most of the plant species suitable for producing this kind of ethanol — like switchgrass, a fast- growing plant found throughout the Great Plains, and farmed poplar trees — aren’t food crops. And according to a joint study by the US Departments of Agriculture and Energy, we can sustainably grow more than 1 billion tons of such biomass on available farmland, using minimal fertilizer. In fact, about two-thirds of what we throw into our landfills today contains cellulose and thus potential fuel. Better still: Cellulosic ethanol yields roughly 80 percent more energy than is required to grow and convert it.
So a wave of public and private funding, bringing newfound optimism, is pouring into research labs. Venture capitalists have invested hundreds of millions of dollars in cellulosic-technology startups. BP has announced that it’s giving $500 million for an Energy Biosciences Institute run by the University of Illinois and UC Berkeley. The Department of Energy pledged $385 million to six companies building cellulosic demonstration plants. In June the DOE added awards for three $125 million bioenergy centers to pursue new research on cellulosic biofuels.
There’s just one catch: No one has yet figured out how to generate energy from plant matter at a competitive price. The result is that no car on the road today uses a drop of cellulosic ethanol.
Cellulose is a tough molecule by design, a fact that dates back 400 million years to when plants made the move from ocean to land and required sturdy cell walls to keep themselves upright and protected against microbes, the elements, and eventually animals. Turning that defensive armor into fuel involves pretreating the plant material with chemicals to strip off cell-wall protections. Then there are two complicated steps: first, introducing enzymes, called cellulases, to break the cellulose down into glucose and xylose; and second, using yeast and other microorganisms to ferment those sugars into ethanol.
The step that has perplexed scientists is the one involving enzymes — proteins that come in an almost infinite variety of three-dimensional structures. They are at work everywhere in living cells, usually speeding up the chemical reactions that break down complex molecules. Because they’re hard to make from scratch, scientists generally extract them from microorganisms that produce them naturally. But the trick is producing the enzymes cheaply enough at an industrial scale and speed.
Today’s cellulases are the enzyme equivalent of vacuum tubes: clunky, slow, and expensive. Now, flush with cash, scientists and companies are racing to develop the cellulosic transistor. Some researchers are trying to build the ultimate microbe in the lab, one that could combine the two key steps of the process. Others are using “directed evolution” and genetic engineering to improve the enzyme-producing microorganisms currently in use. Still others are combing the globe in search of new and better bugs. It’s bio-construction versus bio-tinkering versus bio-prospecting, all with the single goal of creating the perfect enzyme cocktail.
President Bush, for one, seems to believe that the revolution is imminent. “It’s an interesting time, isn’t it,” he mused this February. “We’re on the verge of some breakthroughs that will enable a pile of wood chips to become the raw materials for fuels that will run your car.” Whether the car of the future will be powered by wood chips isn’t clear yet. But it may depend on the success of the hunt for tiny enzymes that could be discovered anywhere from a termite’s stomach in Central America to a lab bench to your own backyard.
Lynd’s microbe would be an all-in-one ethanol factory.
Portrait: Peter Yang
Trace the fortunes of cellulosic ethanol over the past three decades and you’ll find that the arc almost perfectly mirrors Lee Lynd’s career. The 49-year-old Dartmouth professor started in a compost heap in the 1970s, seemed on the verge of a breakthrough in the ’80s, and nearly went bust in the ’90s. “There were times,” he says, “when my lab barely had a pulse.” Now, as a central player in the burgeoning cellulosic industry, he works out of a rejuvenated Dartmouth lab and sparkling new offices in nearby Lebanon, New Hampshire, freshly equipped and staffed by nearly two dozen PhDs. Many are recent hires, the beneficiaries of $60 million that Lynd’s company, Mascoma, has raised. The firm is beginning construction on a pilot-scale ethanol plant in New York state this year, and it recently announced plans for a $100 million production plant in Michigan, projected to break ground in 2008.
Lynd has deep-set eyes and wavy blond hair graying at the temples. He dresses the casual businessman, his inner environmentalist betrayed only by a pair of leather sandals. Working on a farm as a biology undergrad one summer in the 1970s, Lynd noticed that a thermometer stuck in a compost pile registered 150 degrees Fahrenheit. He knew that microorganisms must be at work in there, digesting the plants and turning them into… something. Lynd became obsessed with harnessing that biology to generate usable energy from plants.
He certainly wasn’t the first scientist to try. The oil crisis of the ’70s spurred a wave of federally funded research on cellulosic ethanol. Then, in the mid-’80s, when President Reagan declared the fuel crisis over, the DOE money vanished with few results. Many academics fled to other fields where funding was easier to get. But Lynd — descended from what he calls “several generations of social reformers” — remained enamored with the potential of cellulosic ethanol, and he pieced together small grants to keep his lab running.
For Lynd, the key to the future lies in combining the two main stages of the cellulosic conversion pathway into a single process inside a single microbe. Instead of using enzymes to make sugar out of plant material and then using yeast to convert that sugar to ethanol, Lynd is trying to create a bacterium that serves as an all-in-one fuel factory, taking up cellulose and spitting out ethanol. Called consolidated bioprocessing, or CBP, this has been his dream for two decades. “Almost everybody believes it’s doable,” he says. “People disagree whether it’ll take two years or 20.”
To get there, he needs to engineer cellulase production into a sugar-fermenting microbe like yeast or modify a cellulase-producing organism to make it ferment sugar. With plenty of research money in hand, he’s trying to do both. To accomplish the latter, Lynd and his colleagues are working with a cellulase-producing bacterium called Clostridium thermocellum. “You can isolate this puppy out of garden soil, hot springs, compost heaps, forest floors,” Lynd says. In 2005, the researchers proved that a bug very similar to C. thermocellum could be modified to make ethanol. Their goal is now to modify C. thermocellum to do the same. If he succeeds, Lynd’s analysis shows that CBP — by reducing the raw materials and capital required — could cut overall processing costs twofold, potentially the difference between a profitable ethanol plant and a money pit.
Meanwhile, Mascoma is pushing ahead to build factories that will use commercial cellulase enzymes until the superbug is available. That may not happen immediately, but Lynd is patient, having sought a breakthrough for three decades. “I’m not sure if that makes me inspired or an idiot,” he says. “Probably a little of both.”
Joel Cherry, a molecular biologist
Cherry is making existing enzymes cheaper and more efficient.
Portrait: Peter Yang
If you want to buy enzymes off the shelf, a good place to start would be Novozymes, the world’s leading supplier of cellulases. Headquartered in Denmark, the company runs a tidy business selling millions of pounds of enzymes, used to do everything from brewing alcohol without malt to helping laundry detergent devour stains. Novozymes perfects its enzymes in state-of-the-art biotech labs and sends them to plants scattered around the world, where they are manufactured in bulk. Now, in a subsidiary office tucked away just off I-80 outside Davis, California, the company is prepping its next advance.
Back in 2000, Joel Cherry, a molecular biologist who now runs the company’s research on biomass enzymes, began urging Novozymes to develop some that could be used to produce fuel. “There were a lot of people who said it wasn’t worth doing,” he recalls. But Cherry pressed the company to apply for a DOE grant, and the agency awarded Novozymes and Palo Alto-based Genencor about $15 million each to make the currently available cellulases cheaper and more efficient at chopping up plants. Cherry now heads a team of nearly 100 researchers focused exclusively on cellulosic enzymes, the company’s largest single R&D effort.
The enzymes used today to make cellulosic ethanol come from a microbe that was discovered during World War II, eating away at the tents used by US forces in the South Pacific. It turned out to be a tropical fungus named Trichoderma reesei, which secretes a mixture of more than 50 cellulose-processing enzymes. Researchers have since bred strains of it that can produce the stuff much faster. “It’s definitely the gold standard for cellulase production,” Cherry says, holding up a sample plate covered in the green dust of T. reesei spores.
Novozymes sells T. reesei derived cellulases today, primarily to fabric companies that use them to create the stone-washed look for jeans. But profit margins are fatter on jeans than on commodities like fuel, and the enzymes have remained too expensive to make cellulosic ethanol commercially viable.
So Cherry’s team transplanted four new enzyme-producing genes into the fungus — sequences from other cellulase-generating organisms in the company’s culture collection. For some of the samples, bioengineers used what they call directed evolution: They mutated the genes and then used high-throughput screening to test the resulting enzymes for improvement in properties like heat resistance and ability to degrade cellulose. The best of the mutated-enzyme combinations were then tested in tabletop reactors on corn stover, the cellulose-laden stalks of the crop. After four years, Cherry and his team say they’ve reduced the cost of the enzyme mixture from $5 per gallon of ethanol to well under a dollar. Genencor claims similar improvement.
The only way to truly judge the enzymes’ cost and effectiveness, however, is to put them to work on real feedstocks under industrial conditions. To that end, Novozymes is currently supplying its new enzymes to several companies in the US, Europe, and China that are building cellulosic demonstration plants. Those are among over a dozen outfits — from a company using a thermochemical process to break down wood chips in Georgia to a Massachusetts-based firm that is working on a CBP bug to rival Lynd’s — scrambling for the first commercial cellulosic success.
“We’re at the place now where the enzymes could be significantly cheaper, and we are going to continue to pound on it,” Cherry says. “If one of those efforts can show a clear path to economic viability, I think it’s just going to go crazy.”
John Doyle, Verenium’s vice president
Doyle looks to nature for better enzymes.
Portrait: Peter Yang
Could there be better enzymes in the wild, as yet unknown, just waiting to be discovered? Verenium, based in Cambridge, Massachusetts, thinks so, and it’s prospecting the globe for a bug that produces them. The company’s scientists will go just about anywhere — they’ve explored the excrement of rhinos and the stomachs of cows — but their most intriguing work so far took them to Costa Rica, home to one of the world’s most diverse insect populations. There, working with Caltech microbiologist Jared Leadbetter and a group of Costa Rican scientists, the team gathered termites from the rain forest floor.
Termites are master cellulose processors, using a mixture of bacteria, fungi, and other microorganisms in their hindmost gut to break down leaves and dead trees. “There are lots of organisms that naturally degrade and digest plant cell-wall material,” says biologist Kevin Gray, the company’s director of alternative fuels. “Termites are top on the list.”
After pinching out the termite’s gut, which holds a microliter of material containing an entire ecosystem of microbes, they shipped it back to the US and isolated the DNA. Now, together with the DOE, they’re sequencing that DNA to find the genes responsible for creating the cellulases. A preliminary analysis shows “a large diversity of enzymes,” Gray says. Next, they’ll determine the most effective mix of cellulases by testing what they’ve extracted on plant matter. They hope to find one that will chew up cellulose bonds faster and more efficiently than anything Novozymes’ T. reesei fungus churns out.
But Verenium is not an enzyme company like Novozymes — it’s in the fuel business. Just outside the farm town of Jennings, it also runs a pilot-scale biorefinery amid the steaming bayous of western Louisiana. This is one of the few places in the world where enzymes are already on the job, turning plants into usable fuel.
The process starts with a three-story-high mound of bagasse, a woody byproduct of sugarcane that farmers often discard. The bagasse, which resembles sweet-smelling mulch, travels on a conveyer belt through stainless steel pipes where it’s treated with an acid mixture. Then it’s dumped into 10-foot-diameter tanks for the two biological stages of the process. First, microbes that churn out cellulose-chomping enzymes are funneled into the batch, turning the bagasse into sugar. Then, two micro organisms — including a special strain of Escherichia coli bacteria developed by University of Florida microbiologist Lonnie Ingram — are used to ferment the sugar into alcohol.
This facility churns out enough ethanol to test the basic technology, if not to prove its viability at commercial volumes. But John Doyle, Verenium’s vice president for projects, is overseeing the construction of a larger demonstration plant and hopes to show that the economics can scale, even before the company finds the right termite-derived enzymes.
“The high tech part of our process is the organisms,” Doyle says, “and you can always swap new organisms into the infrastructure.” The refinery, in other words, is just hardware, while the biology supplies the software — with the enzymes upgraded whenever a new, better one is pinched out and perfected.
Skeptics argue that rosy projections for cellulosic ethanol ignore its drawbacks — mainly, that cars need to be converted to run on it, that existing oil pipelines can’t transport it, and that we don’t have the land to grow enough of it. Advocates counter that if the fuel is cheap and plentiful enough, the infrastructure will follow. “If we could make ethanol at a large scale in a way that is sustainable, carbon-neutral, and cost-effective, we would surely be doing so,” Lynd says, citing the fact that most cars can easily be converted to run on ethanol, something already done with most new cars in Brazil. “Meeting these objectives is not limited by the fuel properties of ethanol but rather by the current difficulty of converting cellulosic biomass to sugars.”
Neither government funding nor venture capital, of course, guarantees research breakthroughs or commercial blockbusters. And even ardent proponents concede that cellulosic ethanol won’t solve our fuel problems — or do much to stop global warming — without parallel efforts to improve vehicle efficiency. They also worry that attention could again fade if the first demonstration plants fail or oil prices plummet. “To get this industry going, you need some short-term breakthroughs, by which I mean the next five to seven years,” says Martin Keller, a micro biologist at Oak Ridge National Laboratory in Tennessee and director of its new BioEnergy Science Center. “Otherwise, my fear is that people may leave this field again.”
The problem comes from the quotidian difficulties of making benchtop science work on an industrial scale. Undoubtedly, even some well-funded efforts will fail. But the proliferation of research — the prospect of Lee Lynd’s superbug, the evolution of current cellulases, and the addition of new enzymes harvested from nature — stacks the deck in favor of cellulosic ethanol.
Alexander Karsner, assistant secretary for the DOE’s Office of Energy Efficiency and Renewable Energy, says that with plants going up around the country, the industry could make cellulosic ethanol cost-competitive within six years. “I think there won’t be a silver-bullet process, where you say, ‘That has won, and everything else is done,'” he says. “So you need many of these technologies.”
Having known lean times, Lynd is reluctant to predict the future. But given the freedom of fat wallets, he says, “I truly think that in five years all the hard issues about converting cellulosic biomass to ethanol may be solved.”
The researchers’ vision, of green and gold switchgrass fields feeding a nationwide network of ethanol plants and filling stations, often has an effortless quality to it — as easy as a few steps sketched out on a blackboard. The money and momentum is here. Solve the science, they argue, and the market will take care of the rest.
Contributing editor Evan Ratliff (www.atavistic.org) wrote about Google Maps and Google Earth in issue 15.07.
WIRED MAGAZINE: ISSUE 15.10