WHEREVER automotive engineers gather, some wag will sooner or later announce that hydrogen is the fuel of the future—and always will be. The hydrogen-powered car has been just around the corner for decades. However, judging from announcements by Honda, Hyundai and Toyota at last week’s motor shows in Los Angeles and Tokyo, hydrogen cars will be hitting the showrooms from spring 2014 onwards. It seems the future is about to arrive.Hydrogen’s attraction as a transport fuel is that, unlike petrol, diesel, kerosene, natural gas and every other hydrocarbon fuel, it contains, well, no carbon. Burning it therefore creates no carbon-based greenhouse gases—at least, not in the engine. However, if air is used as the oxidiser instead of pure oxygen, burning hydrogen produces all the noxious oxides of nitrogen that fossil fuels generate. These are an even bigger curse than carbon dioxide as far as damaging greenhouse gases are concerned.That is why work on using hydrogen as a fuel for a modified internal-combustion engine has been more or less abandoned, even though getting such a power unit into production was considered cheaper than any of the clean alternatives. BMW built a couple of hydrogen-powered supercars, only to find them no cleaner than clunkers from the days before catalytic converters.Hence the embrace of fuel cells, which extract chemical energy from hydrogen without resorting to combustion. The process is essentially the opposite of electrolysis: instead of using electricity to split water into hydrogen and oxygen, a fuel cell combines the two gases electrochemically to produce water, while generating an electric current in the process. A fuel cell’s only emissions are thus water vapour and heat.At its simplest, the “PEM” (short for proton-exchange membrane) type of fuel cell used in cars has two electrodes—an anode and a cathode—separated by an electrolyte in the form of a polymer membrane coated with a platinum-palladium catalyst. Hydrogen from a fuel tank is pumped into the anode side of the cell, while the cathode is surrounded by oxygen drawn from the air.As the hydrogen seeks to migrate from the anode to the cathode, the catalyst strips electrons from it, allowing only hydrogen ions (protons) to pass through and migrate to the cathode, where they combine with the oxygen to form water. The electrons rejected by the membrane are diverted via an external circuit to perform useful work (drive a motor, power a heater, light a bulb, etc) before reaching the cathode to complete the circuit. As a PEM cell can deliver a little under one volt, many of them have to be stacked together like a sliced loaf to produce a useful voltage.Fuel-cell stacks are potentially three or four times more efficient than internal combustion engines. More to the point, cars using them are essentially electric vehicles, but without the heavy battery. As such, they solve two big problems that plague battery-powered electric vehicles: their limited range and their long recharging time. Vehicles powered by the latest fuel-cell stacks can achieve over 300 miles (480km) on a tankful of hydrogen. Filling the tank takes five minutes at most. And, like battery electrics, they are classed as zero-emission vehicles (ZEVs).This matters because carmakers have to show state authorities (California’s Air Resources Board, in particular) that they are working hard to meet those states’ zero-emission sales targets. They have grown despondent about battery-powered electric cars—with their paltry ranges and long recharging times—being able to do the job, even if the batteries were to halve in price or double in range.By contrast, hydrogen vehicles—which behave more like conventional cars—could help them get closer to the mandated requirements. By 2025, car companies will need to have sold at least 1.5m zero-emission vehicles in California under the latest clean-air rules. In a normal year, Californians buy around 1.7m new cars. That means something like 15% of all new cars sold in the state will have to be ZEVs by 2025.Seven other states, including New York, Connecticut and Massachusetts, have now adopted similar rules. Between them, the eight states in question account for one out of four new cars bought in America.Another attraction is the rate at which fuel-cell costs are declining. Despite the billions of dollars thrown at lithium-ion technology—the battery of choice for plug-in electrics—progress seems to have stalled, while costs have remained stubbornly high at around $2,000 per kilowatt of power. The contrast with fuel cells could not be greater. When unveiled to the public in 2007, the 100-kilowatt stack in the Honda Clarity hydrogen car was reckoned to have cost the company $350,000 per unit—ie, $3,500 per kilowatt. No wonder Honda built only 200 Clarity test cars for public trials.Over the past six years, however, manufacturers have more than halved the cost of their experimental stacks. Today, they cost less than $1,500 per kilowatt to make in short runs (see “End of the electric car”, October 15th 2012). If they were mass-produced in tens of thousands instead of by the handful, the Department of Energy believes they could be manufactured for less than $50 per kilowatt—much the same as a typical internal combustion engine.The other obstacle to the wide deployment of PEMs—the expense of their catalysts—could be dealt with if a cheap catalyst could be found that worked as well as platinum-palladium, and did not get so easily poisoned by impurities (such as carbon monoxide) in the hydrogen fuel. Some progress has been made with iron-based catalysts, but their activity is still too low to be practical. Alternatively, if the expensive platinum-palladium catalyst could itself be made more catalytically active, then less would be needed to do the job, and fuel cells would inch closer to becoming commercially viable. Given the amount of research underway, both approaches are likely to yield results before the decade is out.All this may not, however, be enough to drag that much-vaunted future into the present. First, there is the problem of providing the hydrogen fuel, along with the infrastructure for transporting it to garages across the country. Of the 100 or so hydrogen-filling stations dotted around America, fewer than a dozen are open to the public. The rest are reserved by industry, the armed services, government agencies and research institutions for their own private use.California’s state government has approved plans to spend $20m a year over the coming decade to build more than 100 hydrogen stations for public use. By one estimate, covering the rest of the country with the barest network of such stations would cost $20 billion. Another study suggests making hydrogen dispensers as common as petrol pumps would cost America the small sum of half a trillion dollars.Then there is the question of where the hydrogen comes from. At present, industrial hydrogen (which is used as a feedstock for refining oil, as well as for making chemicals, electronics and foodstuffs) is produced by reforming natural gas with steam. This is not a particularly clean process. According to the National Renewable Energy Laboratory, a federal facility in Colorado, producing a kilogram of hydrogen by steam reformation generates 11.9 kilograms of carbon dioxide. As the Honda Clarity could travel 68 miles (109km) on a kilogram of hydrogen, it would cause 175 grams of carbon dioxide to be dumped into the atmosphere for every mile it was driven.By way of comparison, Volkswagen’s small diesel cars produce 145 grams per mile. On that reckoning, even petrol-electric hybrids like the Toyota Prius, which produces 167 grams per mile, are cleaner than the fuel-celled Clarity. Admittedly, fossil fuels also produce carbon emissions while being dug out of the ground, refined and transported to the pump. But burning hydrocarbons in internal-combustion engines is becoming cleaner all the time. When measured on a well-to-wheels basis, the steadily declining emission levels of conventional vehicles is putting the squeeze on so-called ZEVs.If the zero-emission rules of California and elsewhere are to mean anything, then the hydrogen has to be made by clean and costly electrolysis of water rather than by cheap and dirty steam reformation of natural gas. And the electricity used to separate water into hydrogen and oxygen has itself to produce no greenhouse gases while being generated. The only way to do that is to rely strictly on renewables like wind and solar energy, or carbon-free hydroelectricity and nuclear power.Unfortunately, renewables do not scale particularly well. Meanwhile, precious few hydroelectric sites are left to be exploited. And nuclear power is, at best, a long-term option these days.Lacking clean electricity, the plug-in electric vehicles and hydrogen cars that manufacturers are being pressed to produce to meet the ZEV goals could wind up being bigger polluters than the petrol and diesel vehicles they replace. That is the message carmakers hope will be heard loud and clear by lawmakers everywhere, for, as they are likely to lose money on every ZEV they sell, they are hoping the authorities will relax the rules once the goals are seen to be unachievable—at least, at prices motorists can afford. Such accommodations have been made several times before. If they are made again, the future of the hydrogen car will stay where it has always been: in the future.