This is what I understand about efficiency after working with photodetectors (not used as solar cells, but similar in principle).
1. The sun emits a wide spectrum, with a long tail in the infrared.
2. Materials are only photoelectric at certain wavelengths.
3. The sensitivity to each wavelength depends upon design which is ultimately is related to the penetration depth of light in the material (longer wavelengths travel longer distances before reacting with the crystal structure).
4. Silicon is the most common and cheapest material, but its sensitivity rolls off sharply towards UV and infrared. Other materials are orders of magnitude more expensive. Not just in manufacturing, but the materials themselves.
5. To capture a wide range of spectrum you need a very long depletion region (kind of neutral area between p-n junctions). This has its own design tradeoffs.
So the end result is that the real world efficiency limit can have many reported figures, from theoretical energy physics to practical material, design, and manufacturing limits. The biggest contributor to loss is that silicon is simply not able to capture much more than the visible spectrum. Next is that even with silicon there is a sensitivity curve (efficiency of converting photons to current per wavelength), and then designing so that you can capture the current (or else it would just recombine in the crystal).
This is why gain in solar seems to be slow. There are tremendous limitations, and I have my doubts that there will be overcome any time soon.
But, there is plenty of sunlight, so I would say the real challenge is energy storage. If sunlight can be stored cheaply and efficiently, then what we have is good enough.
I often wonder how good/cost effective optics would be, given the price of glass is cheap and that could be used to focus a larger area upon a small, more costly and yet efficient material.
But as you highlight, cost and with that a balance of cost/return is a huge factor and whilst scale can help in many area's. It gets more detrimental with more expensive materials.
I don't know the exact answer, but one should be able to calculate how effective this method would be. You are knocking out electrons from a crystal. Those electrons have to be replaced, so you are limited by the rate of replacement. Also heat generation from the current.
My hand wavy gut feel is that optics wouldn't buy you that much. Silicon is orders of magnitude cheaper. To give you an idea, an infrared detector made of InGaAs about the size of a dime can already cost $5-10k. Solar cells of that size are priced in ... dollars?
On the other hand, optics have been used for energy/storage in the form of a heating crucible. I've heard that promoted by some experts.
1. The sun emits a wide spectrum, with a long tail in the infrared.
2. Materials are only photoelectric at certain wavelengths.
3. The sensitivity to each wavelength depends upon design which is ultimately is related to the penetration depth of light in the material (longer wavelengths travel longer distances before reacting with the crystal structure).
4. Silicon is the most common and cheapest material, but its sensitivity rolls off sharply towards UV and infrared. Other materials are orders of magnitude more expensive. Not just in manufacturing, but the materials themselves.
5. To capture a wide range of spectrum you need a very long depletion region (kind of neutral area between p-n junctions). This has its own design tradeoffs.
So the end result is that the real world efficiency limit can have many reported figures, from theoretical energy physics to practical material, design, and manufacturing limits. The biggest contributor to loss is that silicon is simply not able to capture much more than the visible spectrum. Next is that even with silicon there is a sensitivity curve (efficiency of converting photons to current per wavelength), and then designing so that you can capture the current (or else it would just recombine in the crystal).
This is why gain in solar seems to be slow. There are tremendous limitations, and I have my doubts that there will be overcome any time soon.
But, there is plenty of sunlight, so I would say the real challenge is energy storage. If sunlight can be stored cheaply and efficiently, then what we have is good enough.