In the 1990s a graduate student named Lin Chao at Princeton University decided to bubble carbon dioxide into an electrochemical cell. Using cathodes made from the element palladium and a catalyst known as pyridinium—a garden variety organic chemical that is a by-product of oil refining—he discovered that applying an electric current would assemble methanol from the CO2. He published his findings in 1994—and no one cared.

But by 2003, Chao's successor in the Princeton lab of chemist Andrew Bocarsly was deeply interested in finding a solution to the growing problem of the CO2 pollution causing global climate change. Graduate student Emily Barton picked up where he left off and, using an electrochemical cell that employs a semiconducting material used in photovoltaic solar cells for one of its electrodes, succeeded in tapping sunlight to transform CO2 into the basic fuel.

"The dominant thinking 10 years ago was that we should bury the CO2. But if you could efficiently convert it into something that we wouldn't have to spend all that money and energy to put into the ground, sort of recycle it, that would be better," Bocarsly says. "We take CO2, water, sunlight and an appropriate catalyst and generate an alcoholic fuel."

He adds: "We didn't have some brilliant insight here. We had some luck." Luck that venture capitalists are now trying to turn into cash flow via a start-up known as Liquid Light.

Turning CO2 into fuels is exactly what photosynthetic organisms have been doing for billions of years, although their fuels tend to be foods, like sugars. Now humans are trying to store the energy in sunlight by making a liquid fuel from CO2 and hydrogen—a prospect that could recycle CO2 emissions and slow down the rapid buildup of such greenhouse gases in the atmosphere. "You take electricity and combine CO2 with hydrogen to make gasoline," explained Arun Majumdar, director of the Advanced Research Projects Agency–Energy (ARPA–e) that is pursuing such technology, at a conference in March. "This is like killing four birds with one stone"—namely, energy security, climate change, the federal deficit and, potentially, unemployment.

"When these new technologies get commercialized, those jobs always end up in the U.S.," argues chemical engineer Alan Weimer of the University of Colorado at Boulder, who is working on such solar-fuel generators. Adds chemist Michael Berman of the U.S. Air Force Office of Scientific Research, which is funding research into the possibilities of solar fuels, including Bocarsly's work: "The country, and the Air Force, need secure and sustainable sources of energy…. Since the sun provides enough energy for our needs, our goal is to make a fuel using CO2 and sunlight—and maybe water—as feedstocks to produce the chemical fuel that can store the sun's energy in a form that we can use where and when we need."

In fact, the problem with turning CO2 back into a hydrocarbon fuel is not so much in transformation—there are at least three potential approaches to do so with sunlight, along with a process that employs high pressures and temperatures, so-called Fischer-Tropsch, which is currently used—but rather the tremendous expense involved. "It's an uphill process to convert CO2 to methanol. It's going to cost you some energy," Bocarsly says. "The current rate of generating methanol is not high enough to be commercial."

Solar heat
The sun bathes Earth in more energy in an hour than human civilization uses in a year. Giant dish mirrors in the New Mexico desert erected by scientists at Sandia National Laboratories capture some of that energy and have been used to concentrate it on a cylindrical machine that looks like a beer keg—a would-be solar-fuel generator. Nestled inside that machine are a series of rotating, concentric rings. Turning at roughly one rotation per minute, the CR5—for counter rotating ring receiver reactor recuperator—moves these rings studded with teeth containing iron oxide (also known as ferrite, or rust) or cerium oxide (ceria) into and out of the sunlight. The sun heats the teeth as high as 1,500 degrees Celsius—driving oxygen out of the rust—before they rotate back into the darkness and cool off to roughly 900 degrees C. In the darkness, steam or CO2 is injected and the greedy ferrite sucks the oxygen out of those molecules—leaving carbon monoxide (CO) or hydrogen (H2) behind—before rotating back into the sun.

The resulting CO–H2 mixture is so-called synthesis gas, or syngas—the basic molecular building block of fossil fuels, chemicals, even plastics. The CR5 "is a chemical heat engine," says chemical physicist Ellen Stechel, program manager of Sandia's Sunshine to Petrol project that basically seeks to reverse fossil-fuel combustion. "It's doing chemical work, breaking a bond."

The CR5 has already been turned on three times—and will likely be run again this year before the sun gets too weak, according to chemical engineer James Miller of Sandia, co-inventor of the device. But it has never reached the steady state necessary to efficiently throw off syngas. The problem is the thousands of ceramic tiles that form the reactive teeth on the edge of the rotating rings, some of which break as the process heats up. "You're cycling back and forth from 1,500 to 900 [degrees C], and that's a lot to ask of a material," notes chemist Gary Dirks, director of LightWorks at Arizona State University, who is not involved with the project.

Other groups are working on different designs or different materials, such as zinc oxide, but finding better materials is only one part of the challenge—the Sandia researchers estimate they could make a precisely equivalent replacement gasoline, diesel or jet fuel for roughly $10 per gallon. Simply put, all those specialty mirrors to concentrate the sunlight and metal structures to hold those mirrors in place are expensive. "Even though sunlight is free, what costs you the most is collecting it and converting it into a useable form," Miller says. "Sunlight is the make-or-break feedstock."

But there is another problem: to replace the 20 million barrels of oil used every day in the U.S. alone—roughly 60 percent of which is imported—would require roughly 685 million concentrating solar dishes covering more than six million hectares of the desert Southwest. It would also require 62.4 trillion moles of concentrated CO2 per year. "That's the miracle we haven't addressed," Miller says.

Even capturing and concentrating the emissions from the nation's fleet of fossil fuel–fired power plants—which would add even more to the expense of the process—would not be enough. "Liquid fuel to burn in your car is the ultimate manifestation of making things at scale," notes chemist Eric Toone, an ARPA-e program manager working on so-called "electrofuels," or hydrocarbons manufactured via microbes. "When you think of how much oil we actually need, that source of CO2 becomes a really great question."

Artificial photosynthesis
Nature has an answer. Plants pull CO2 out of the air and, thanks to the ongoing burning of the results of millions of years of photosynthesis (otherwise known as fossil fuels), atmospheric concentrations continue to rise. Unfortunately, plants are woefully inefficient at turning sunlight into food—averaging at best 1 percent of the incoming sunlight stored as chemical energy, thanks to competing concerns like survival—one main reason that the U.S. Department of Energy (DoE) estimates that at best, 15 percent of the nation's energy needs could come from biofuels (pdf).

Chemist Nathan Lewis of the California Institute of Technology would like to improve on that by mimicking the processes of photosynthesis—light absorbers, molecule-makers and membranes to separate various products, among other things—artificially. "Nature uses enzymes; we use inorganic complexes or metals," Lewis says of a new effort to create artificial photosynthesis launched with DoE funding on July 22. "Materials carry as much [electric] current as you like because they move electrons rather than molecules."

Lewis notes that all of these artificial processes exist on their own—but do not necessarily work well together. The goal over the next five years of the Joint Center for Artificial Photosynthesis, a project that Lewis directs, will be to simply prove that it can be done. "If we demonstrate that we can make solar-fuel generators, it would be like the Wright brothers," he says. "It's not a 747, but it shows humans can fly."

Lewis is not alone. Massachusetts Institute of Technology chemist Dan Nocera is working with novel catalysts to improve water-splitting—the vital step for deriving the hydrogen that is then paired with CO2 to make hydrocarbons. NASA has funded scientists to research turning CO2 to fuels in order to make it possible for Mars explorers to manufacture rocket fuel from the Martian atmosphere for their return trip to Earth. And Mantra Energy has paired with utility KOSPO to employ its electrochemical cell technology to convert CO2 to formic acid—an essential building block for many chemicals or fuels— at one of the company's coal-fired power plants in South Korea. "What they're doing with [CO2] now is they are having to release it to the atmosphere" despite having attached dry-adsorption technology to capture the CO2, explains John Russell, Mantra's chief technology officer, although he admits his start-up has yet to build a power plant–size unit. "That is going to take the best part of six months and then we'll run it for a similar sort of time."

Liquid light
The utility of liquid fuels is clear: one gallon of gasoline contains as much energy as 55,000 gallons of water pumped uphill to the height of Hoover Dam and then dropped back through turbines, and the best batteries offer 200 watt-hours per kilogram of energy whereas gasoline delivers 140,000 watt-hours per liter, according to Caltech's Lewis. "There is nothing that can come close to the gravimetric and volumetric density of liquid fuel," ARPA–e's Toone notes. "It's hard to see how you electrify long-distance trucking and impossible to see how you electrify long-distance flight."

In fact, replacing one oil field that produces 500,000 barrels a day would require a 100-square-kilometer algae-biofuel field, according to former oil man Dirks—and replacing the world's oil habit in that way would require thousands of such fields. "It will take at least 20 years before we get a material change in the existing system, by which I mean 15 to 20 percent of our liquid fuels coming from something we don't do right now," he says. "That's as fast as we can do it."

In the last 150 years or so humanity has already run through a few hundred million years of ancient photosynthesis conveniently stored underground. Whereas vast reserves of such fossil sunshine remains—think of the tar sands in Canada or coal beds in Siberia—the cost of utilizing them is an entirely altered climate from the one that has allowed human civilization to flourish. Plus, "this is a finite resource," Toone notes.

Liquid Light would like to push back that deadline by allowing CO2 molecules to be recycled via a version of the electrochemical cell from Bocarsly's Princeton lab. "If you can get even one more cycle out of the same CO2 molecule before you put it in the atmosphere, it's much more efficient," says physicist Nety Krishna, CEO for Liquid Light. "Essentially, you delay the increase in CO2 in the atmosphere."

The company has replaced the expensive platinum electrode in the original cell and has eliminated the semiconducting electrode entirely. "In the beginning we will use electricity as the source of electrons," Krishna admits. They don't even use pyridinium anymore, although it is cheap and abundant, and Bocarsly reports being able to make not just methanol—not ideal as a fuel—but also longer-chain hydrocarbons with it. For the moment Liquid Light plans to pursue the manufacture of more valuable chemicals or simply syngas alone before chasing the dream of a drop-in replacement for fossil fuels, which would require scaling up from the company's liter-scale prototype to one with hundreds of millions of liters of capacity. "That's the scale in order to make any difference," Krishna notes.

The company hopes to have a prototype system built in the next year or so. The CR5 will continue to be tweaked and improved, and alternate designs are in the works at the University of Minnesota and elsewhere. And, already, companies like Sundrop Fuels are using the sun's heat to gasify the plant stalks and other biomass built by photosynthesis to make a better fuel. But it will take more than a cost to emitting CO2 in the U.S. to drive the development of this technology, says chemical engineer Jane Davidson of the University of Minnesota. "It requires an economic stimulus."

But the promise is an abundance of liquid fuels that store sunlight as chemical energy and are slower to exacerbate atmospheric concentrations of greenhouse gases. "As taken from the example of nature, if you can store [sunlight] in chemical bonds then you solve the intermittency problem, because the sun has this nasty habit: it goes out locally every single night," Caltech's Lewis says. "Chemical fuels would be the game changer if you could directly make them efficiently and cheaply from sunlight. It's pairing the biggest source [of energy] and the biggest storage."