Unless this work differs completely from Nocera's previous work, the device relies on a conventional solar cell (with sub 20% efficiency if made from silicon as

Unless this work differs completely from Nocera's previous work, the device relies on a conventional solar cell (with sub 20% efficiency if made from silicon as suggested in the press release) to provide current for electrolysis of water. Nocera's group focuses on developing low-cost, durable catalysts for electrolysis that can work efficiently at neutral pH. This would allow electrical energy (from solar or other sources) to be stored as hydrogen gas. However, there are other ways to store solar energy once it is converted to electrical energy or chemical energy. The bigger impediment to large scale adoption of solar is the high cost and low efficiency of initially capturing the energy in each photon. Unfortunately, Nocera's interesting work does not address this much greater challenge. Also, why was the device tested for only 45 hours? It would have taken more than two days just to prepare the talk he gave. This almost certainly means that the efficiency started dropping and the experiment was stopped. Anyone interested in this topic should check out the work of Michael Gratzel and others on photoelectrochemical cells, which are more deserving of the label "artificial leaf."