A start-up company has broken ground on a Texas pilot plant that is supposed to produce ethanol and diesel in a radical new way: with an organism that sweats fuel.

The company, Joule Unlimited of Cambridge, Mass., has developed several patented gene-altered organisms that absorb sunlight and carbon dioxide and combine these into hydrocarbons.

The organisms – basically single-celled plants – live in a panel that vaguely resembles a solar photovoltaic one. They lie under a glass sheet that is mounted on a frame to face the sun.

They live in brackish water and need small amounts of chemical nutrients, said William J. Sims, president and chief executive of Joule. A move from the lab to the field will test their tolerance for temperature variations and other natural challenges, he said. They can survive cold, but the site, in Leander, about 30 miles north of Austin, was chosen because ice was unlikely, he said.

Many companies are pursuing the development of fuel from algae that are immersed in water enriched with carbon dioxide and then soak up sunlight. The algae produce fats and lipids internally; the trick is then to crush them and harvest the oil in a way that does not use much energy. That step has not been widely commercialized. 

But Joule says its organisms release their oil and survive to make more. And the diesel fuel is easy to gather because, like most hydrocarbon oils, it is lighter than water and tends to separate. Ethanol mixes with water and must be distilled, but the technology for this is widely available.

Although the technology is still at the laboratory stage, Joule’s system was recently identified by Technology Review, the M.I.T. magazine, as one of the 10 most important emerging technologies.

The Leander pilot is supposed to begin in operation in June. Mr. Sims was guarded in addressing details about the organisms. He said that they were based on what occurred in nature that and each type would secrete a different chemical. The cells will eat, reproduce and then sweat; water flowing through the system will carry off the hydrocarbon fuel for separation, he said.

After about eight weeks, workers will flush the system clean and start over with a new batch, he said. Carbon dioxide is trucked in for now, but the longer-term strategy is to locate the operation near a power plant that runs on coal or natural gas and captures its carbon dioxide. If a national cap on emissions is enacted,  a power plant might be willing to pay a fuel plant to take its carbon dioxide gas.  Another potential source is a conventional corn ethanol plant, which produces a pure stream of carbon dioxide.

Last week Joule said it had raised $30 million in a second round of financing. The scale-up from the demonstration plant is simple, Mr. Sims said: just reproduce the cluster of panels, the piping that connects them and the separation plant.

The company projects production of 25,000 gallons of ethanol a year from each acre, which would be many times higher than production from wood waste or other biomass source.

Of course all of this remains to be demonstrated, but Joule hopes to be in commercial production in 2012.