In such a system, the carbon dioxide released by burning methanol would be cancelled out by the carbon dioxide captured to make it. So the process would be carbon neutral, and the methanol produced would be a convenient liquid fuel that could replace petroleum-based fuels. If the carbon dioxide comes from air and the hydrogen from water, this method of making methanol would be like fast photosynthesis: "We don't have to wait for plant life to slowly convert excess carbon dioxide into hydrocarbons," Olah says. "We can substitute for Mother Nature."

Olah emphasizes that the methanol produced in this way would not be a new energy source, but simply a convenient way of storing energy. Its advantage over hydrogen would be the ability to use existing engines and infrastructure with only minor modifications.

In many ways, with its low emissions and an octane rating of 100, methanol is already a better fuel for internal combustion engines than gasoline. A methanol engine can run at a higher compression ratio, and is easier to cool. But methanol has some drawbacks: it has lower vapor pressure than gasoline, which makes engines sluggish on cold starts, and it burns with an invisible flame, which could be a safety hazard, since it would be hard for emergency workers to detect in an accident, for example. To mitigate these problems, methanol today is usually blended with 15 percent gasoline to make a fuel mix known as "M85."

Methanol is an even better automotive fuel when used in combination with fuel-cell technology, says Paul Erickson, assistant professor in mechanical engineering at the University of California, Davis. Fuel cells, which convert chemical energy directly into electricity, are more efficient than engines that burn fuel. The hydrogen fuel cell, in particular, has been widely proposed as a clean and efficient alternative to gasoline-powered internal combustion engines. Erickson's laboratory has a functioning hydrogen fuel-cell bus with an onboard reactor that "reforms" methanol to produce hydrogen for its fuel cells. "We completely avoid having to store hydrogen," Erickson says.

Onboard "reforming," however, consumes space and energy. In 1993, Prakash, Olah, and a team at the Jet Propulsion Laboratory in Pasadena, CA, jointly invented a fuel cell that runs directly on a mixture of methanol and water. The cell's positive and negative electrodes are separated by a membrane designed to allow only protons from the methanol to migrate from one electrode to the other. Early versions of this membrane, however, allowed some methanol to get across and react with oxygen at the second electrode, which reduced the voltage of the cell and wasted energy in the form of heat.

In 2001, Prakash and his colleagues developed a new membrane that is both cheaper and more resistant to crossover. With this refinement, the direct methanol fuel cell gives an efficiency of 35 percent, about twice that of an internal combustion engine, but well short of its theoretical efficiency of 97 percent.

The direct methanol fuel cell is currently too expensive to be used in passenger cars. Its high cost comes mainly from the platinum and ruthenium used as catalysts. Prakash and others are developing a variety of approaches to reduce the amount of catalyst needed: making the catalyst more active, increasing its surface area, and using nanoscale methods. When this technology matures, Erickson believes it might replace the hydrogen fuel cell. "An inexpensive, high-power direct methanol fuel cell is the Holy Grail," he says.