You can power laptops - and, potentially cars - using hydrogen extracted from water. The trouble is that it takes a lot of electricity. A simpler way would be to do it naturally, using enzymes - proteins which catalyse reactions - and bacteria. These do exist: certain green algae and "cyanobacteria" can split water using photosynthesis to produce molecular hydrogen.
But to create a generation of cars that would run on water with some sludge in the back, we need to learn how to design our own bacteria and enzymes that can co-opt natural processes for our ends.
Natural hydrogen-producing enzymes are complex, often using metal atoms to help them work. "For many of the enzymes related to energy production, people have no idea how they are actually organised," says Giovanna Ghirlanda, a protein-design researcher at the University of Arizona. In some cases, no one knows where the metal atoms lie within the protein, she says.
Natural enzymes won't work too well in future fuel cells; they need to be modified, as the best hydrogen producers are poisoned by oxygen. "But oxygen is one of the main products of photosynthesis," says Professor Alfonso Jaramillo of the Ecole Polytechnic, near Paris.
Some researchers are trying to tweak the enzymes to make them less sensitive to oxygen, but with limited success. As a part of the EU-funded BioModularH2 project, Jaramillo's team is using a different approach: stick with the natural enzyme and engineer another set of proteins that take oxygen out of the cell before it can do any harm. These hydrogen producers are longer-term options: it may take 10 years to get to a prototype, says Jaramillo.
Tough chemistry
But tweaking enzymes and bacteria to our own ends - a process called "synthetic biology" - is potentially big business. For instance, scientists are working on a new generation of corn that will rot from the inside out once harvested and heated, to produce higher-yielding biofuels.
In the US, corn kernels provide the main source of ethanol for fuel. But typically half of the plant's mass winds up in the tough leaves and stalks; farmers discard this as "stover". The cellulose inside stover is a rich source of sugars that can be fermented into ethanol, but breaking that down is hard and expensive.
"You have to use tough chemistry to do this," says Professor Arthur Ragauskas of the Institute of Paper Science and Technology at the Georgia Institute of Technology. The pre-treatment process involves high pressure, temperatures of up to 200C and strong acids. Scientists call the tough mixture of cellulose and woody lignins in crops "recalcitrance".
"To me, the holy grail is to engineer plants that do not need this kind of pre-treatment," Ragauskas says.
Genetic engineering to make the cell wall softer by taking out the genes that drive the production of lignins wouldn't work either, he says: "It would be a field day for bugs and pests."
Now the US-based company Agrivida aims to insert a gene that will make an enzyme that does nothing until the corn is heated in a reactor to around 60C. Triggered by the heat, the enzyme will change shape and function, attacking the tough materials that form the cell wall.
Artificial enzymes are already used in medicine and biological washing powder. But most are the result of accelerated "directed" evolution from enzymes found in the wild. For its protein switch, Agrivida needed an enzyme designed almost from scratch, as nothing in nature came close to doing the job.
"There was nothing to do direction evolution on. We had to ask: what would create that activity? Once you have the starting point, then you can use directed-evolution techniques," says Brian Baynes, president of Codon Devices, which is producing the genetic material that will go into Agrivida's breed of corn.
Agrivida is one of a number of companies using designer proteins to help produce biofuel. Companies such as Amyris Biotechnologies, Codexis, Gevo, LS9, Mascoma and SunEthanol are trying to develop custom enzymes using synthetic biology to convert non-food biomass into fuel. But unlike Agrivida, most are focusing on growing microbes that will digest the crop in a refinery.
Sunlight solution
Although he acknowledges the regulatory and public-acceptance hurdles involved with producing genetically modified crops, Michael Raab, chief executive of Agrivida, claims: "We think this kind of technology is critical to making affordable biofuels. Whether it is us or someone else that does it, it will be used."
Scientists are still some way from being able to design an enzyme on the computer, insert the gene for it into a plant or microbe and expect it to work. But the ultimate aim of schemes such as BioModularH2 is to go beyond ethanol and use sunlight to power the production of hydrogen from water. Perhaps only then the arguments over biofuels - frequently criticised for taking up land that could be used to grow food - will be settled.
In the meantime, cellulose-derived fuels will be first to benefit from designer proteins. But there remains a question mark over the sustainability of using biomass for fuel, no matter how efficiently it can be produced. "They are using so-called agricultural waste. From a sustainability point of view, there is no such thing: it gets ploughed back into the land," says Jim Thomas, research programme manager at the Canada-based ETC Group, which is pushing for a moratorium on the commercialisation of synthetic biology.
Raab agrees that not all the corn stover can be used for fuel: he estimates that half is needed by farmers to maintain soil fertility. Even so, Raab claims, "instead of making 200 gallons per acre of ethanol now, we could produce 300 gallons".
Ragauskas says the quantity of biomass that needs to be left in the ground is still the subject of research, and it will differ geographically. "Undoubtedly, those solutions will have to be tailored to different regions of each nation, based on climate, soil, growing conditions and the rest. It doesn't mean something tailored for the south-eastern US will be the gold standard for the prairies of Canada," he says.
Even with an agricultural system tuned to extract as much biomass as possible, there will not be enough land to supply all fuel. "We have never said that we are just going to replace all petroleum... you won't solve society's problems with renewable fuels," says Ragauskas. Clearly, our need for designer enzymes is growing urgent.
