Every biofuel also consumes crops that could be feeding a hungry globe. A recent UN report concludes that although the potential benefits are large, the biofuels boom could reduce food security and drive up food prices in a world where 25,000 people die of hunger every day, most under age five. Demand for both fuel and food is expected to more than double by mid-century, and many scientists fear that in coming decades, climate change will undermine agricultural productivity. "Agriculture should be used to stop the hunger of the people. If one person were hungry, this would be a shame," says Goulart. "There are millions who are hungry in Brazil, and this monoculture does not help."

The only way to reap the benefits of biofuels without squeezing the food supply is to take food out of the picture. Though corn kernels and cane juice are the traditional sources of ethanol, you can also make it from stalks, leaves, and even sawdust—plant by-products that are normally dumped, burned, or plowed back under. These materials are mostly cellulose, the tough chains of sugar molecules that make up plant cell walls. Breaking up those chains and fermenting the sugars could yield a cornucopia of biofuels, without competing with food crops. Biofuel visionaries picture a resurgence of deep-rooted perennial prairie grasses like switchgrass or buffalo grass, sequestering carbon in the soil, providing wildlife habitat and erosion control, and supplying a bounty of homegrown fuel.

The principle behind cellulosic ethanol is simple. Making it as cheap as gas isn't.

So far, only a few pilot plants are making ethanol from cellulose in the U.S. A small operation at the National Renewable Energy Lab (NREL) in Golden, Colorado, has been running the longest. It can convert a ton of biomass—shredded cornstalks, switchgrass, wood—into 70 gallons (265 liters) of ethanol in about a week. Along with cellulose and hemicellulose, these feedstocks all contain a substance called lignin. Lignin binds the cellulose molecules together, giving plants the structural strength to stand up and catch the sun. The gluey lignin also makes plant matter hard to break down, as the pulp and paper industry is well aware. "The old joke is you can make anything from lignin but money," says Andy Aden, a senior researcher on the ethanol project.

To unlock the cellulose molecules from the lignin, the feedstock is often pretreated with heat and acid. Then it's mixed with high-tech enzymes to break down the cellulose into sugars. The resulting dark brown goo, with a slightly sweet, molasses-like aroma, is fed into fermentation tanks where bacteria or yeast go to work to make the alcohol. The current process turns just 45 percent of the energy content in the biomass into alcohol, compared with an oil refinery, which extracts 85 percent of the energy in crude oil. The efficiency will have to improve for cellulosic ethanol to compete with gasoline, and researchers are looking for better cellulose-busters. One possibility: genetically modified microbes and enzymes from the guts of termites—nature's own cellulosic energy factories.

The potential, however, is huge. Exploiting the cellulose in corn plants, rather than just the kernels, could double corn's ethanol yield; switchgrass could produce as much ethanol per acre as sugarcane. A 2005 study by the U.S. Department of Agriculture and the U.S. Department of Energy estimated that by boosting farm productivity and planting 50 million acres (20 million hectares) of fallow land with perennial grasses and fast-growing trees, the U.S. could produce 1.3 billion tons (about 1.2 billion metric tons) of feedstock for ethanol. Separately, NREL calculated that all that plant matter could replace more than half the transportation fuel currently burned each year. Mike Pacheco, former director of NREL's Bio-energy Center, pulls out a chart from that study. "The green line is what we think we can make on farms and from trees and switchgrass"—the equivalent of 3.5 billion barrels of oil.

Pacheco traces another line on his chart, at twice the altitude of the first. It represents the ultimate biofuels dream: enough green fuel to make the U.S. energy independent. It is where we might be, says Pacheco, if we greatly increase vehicle efficiency while churning out cellulosic ethanol, or, more tantalizing, "if we make algae work."