The article goes on about the idea of collecting hydrogen from the atmosphere using a graphene filter. There is no appreciable amount of hydrogen in the atmosphere.
At about 140 kJ/kg, and assuming a 10% reduction efficiency, and 100 watts per square meter solar panel, it should be able to reduce 0,071 grams per second or 260 grams per hour.
Assuming 10% of air moisture can be extracted, the wind passing through the system to keep the water flow would need to be about 0,7 m/s. It's almost always as windy as that.
The rough numbers are frighteningly good.
All we need is the system to be cheap.
But I realized that is roughly the momentary day power of the panel. The real yearly average production is maybe 15% of that.
Forgot the panel efficiency with this http://en.wikipedia.org/wiki/Photovoltaic_system#mediaviewer...