There are a number of methods that are presently used to release hydrogen from a number of compounds: electrolysis, steam reforming, and chemical reaction. Each of these has its own advantages and disadvantages.
Electrolysis requires large amounts of electricity and usually some kind of acid to dissociate the oxygen and hydrogen in water.
Steam reforming usually required high temperatures, a catalyst, a source of hydrocarbons (usually some form of petroleum fuel) and some kind of filtration system to separate the hydrogen from the carbon dioxide/monoxide, and other byproducts of the reaction.
Chemical reactions can usually generate a large amount of hydrogen for short periods of time unless the chemicals involved in the reaction are constantly replenished. This kind of reaction can also generate large amounts of heat that need to be disposed of or require large amounts of heat in order for the reaction to even take place.
The problem with all of these methods is that they return only a small fraction of the energy put into them in the form of usable hydrogen. They aren't nearly as efficient as they might be. But that may soon change if researchers at Purdue University have their way.
The method eliminates the need to store or transport hydrogen — two major hurdles on the road to a hydrogen economy, says Jerry Woodall, professor of electrical and computer engineering at Purdue and inventor of the process.
Hydrogen is generated when water is added to pellets of aluminum alloyed with gallium. The aluminum reacts because it has a strong attraction to oxygen in the water. The reaction splits the oxygen and hydrogen, releasing hydrogen in the process.
Gallium is critical because it prevents the skin that normally forms on the aluminum's surface after oxidation. Without the skin, the reaction continues until the aluminum is used up.
I have no idea whether the reaction is endothermic (requires heat to take place), or exothermic (gives off heat during the reaction), but it seems that either way it may be a far more efficient way to generate hydrogen on demand. If it requires a heat source, it might be possible to use solar energy to drive the reaction. Waste heat from industrial processes or power plants could also be used to drive the reaction. If it gives off heat then the waste heat could be used to heat buildings, water, or to drive industrial processes that require heat.
I get the impression that the amount of heat required or generated isn't nearly as much as that of a straight chemical reaction. Since it appears to be a catalytic reaction the only other thing that would be needed to sustain the reaction is a supply of water.
If this is a practical means of generating hydrogen at a reasonable cost, then our changeover to a hydrogen economy may have just come one step closer to reality.