How does a solar cell work?

Paul Alivisatos: How do we look, not from an economic perspective or a market perspective, but just from the point of view of the fundamental physics of how devices operate, how has solar technology evolved. And so in that context we would say the simplest thing that we can have is a device which takes the energy of the sunlight and converts it to heat. And remember that the sunlight consists of these quantum particles that have a couple of electron volts energy each, and those are impinging on a surface and when it becomes hot, we thermalize those photons down to a few -- maybe, 20 millivolts or so. So, the next type of device is the one which is based on the basic photoelectric effect discovered by Einstein and which involves promoting an electron across the band gap of a semiconductor material so that if the photon is bigger than the band gap, electron gets excited and now thermalizes again, but only down to the band gap, and not any further, so we get to keep one whole electron volt, roughly, of energy in that. And in that context, when classified in that way, we'd consider crystalline and thin-film to belong to the same family. Of course, there are multi-gap cells where we separately collect the red, green, and blue photons. And today what a lot of scientists are working on is what is thermodynamically, still further, more advanced technology, which is to make solar fuel. That would be to copy what is done by plants, which take CO2 and water, and then use the energy of photons to make new molecules, to make fuel molecules. If we could make a solar fuel generator with reasonable efficiency, that would thermodynamically be uphill quite a ways because now instead of just capturing the energy of the sun in terms of storing it in the carriers, which we can make electricity from, we would be using that energy to rearrange the atoms within molecules to create a more complex matter, so we'd be going entropically uphill. And so from a long-term perspective of what is thermodynamically the most advanced version of solar energy that would be a higher order version of things, which we would certainly like to get to. And we're also very interested, of course, in thinking about the problem from a top down perspective. In other words, not necessarily looking at where is the PV industry today, but looking a little bit, as much as we can at the problem from the opposite end and saying, where would we like it to be at the end. So, if we look at -- let's say this plot shows the total power consumption of the United States, which is around 3.2, 3.3 terawatts, so it's a large number. Remember global PV production might be 7 or 8 gigawatts in a good year. If we're at 3.3 terrawatts, that's pretty far off, is where we'd like to end up perhaps some day. So, this plot just says let's take a certain area of the United States and as the previous speaker said, we're fortunate in the US to have a lot of land, and so it's possible to imagine doing something like this. It wouldn't happen every place in the world, but here it might be. Let's take a number of 60 million acres, that's about a quarter of the agricultural land, and we imagine converting the energy of the incident photons to a usable form at different power efficiencies shown up there, 1, 3, and 10 percent. And what you can see is that at 1 percent power efficiency, that's a very, very low power efficiency, but if we had that amount of area, 60 million acres that would be equal to US gasoline consumption. And at -- thank you. The pointer, thank you very much. And at 3 -- at 6 or 7 percent, it's really equal to total US energy use. Just to calibrate, a hundred million acres is a quarter -- is 0.4 million square kilometers. And California is about 400,000 acres, roughly.

==== Transcribed by Automatic Sync Technologies ====