Solar Approaches

I recently mused about what would happen if there was an order-of-magnitude improvement in the price per watt of photovoltaic power. I also speculated that there's probably a 25% chance that we might see such a breakthrough in the next decade.

A 25% chance of an order of magnitude improvement in any technology over ten years is pretty aggressive for anything other than disk storage. So why am I so optimistic? Because there seem to be so many promising technologies out there, and there's no particular law of physics which says it can't be done.

To make this more tangible, think of a sheet of ply wood, 4' by 8'. Now imagine that this sheet of plywood is a solar panel. Now imagine that 4x8 solar panel costs $125 installed.

That's what an order of magnitude improvement in the price of solar power means.

That price isn't physically impossible. There are lots of materials out there which can be nailed to a roof for less than $125 for a 4x8 sheet, especially if this solar panel is replacing something else which would otherwise be nailed to the roof (such as a piece of plywood sheathing). It implies not only breakthroughs in photovoltaic technology, but also dramatically simplified installation techniques: the installers would have to pretty much nail down the solar panel, attach a couple wires, and be done.

Several different approaches are currently somewhere between the lab and the factory:

Thin Films

Thin films use technology similar to current silicon solar cells, but instead of manufacturing the solar cell on a silicon wafer, thin film technology deposits the semiconductor in an ultrathin layer on some substrate material. Some approaches use silicon, and others use more exotic semiconductors which outperform silicon in a thin film.

Thin film technologies have two advantages: less material and cheaper manufacturing. Less material is obvious, and that's important today when one of the main drivers of the price of solar cells is the price of silicon. Thin films can use 1% or less of the material as a comparable wafer.

Manufacturing is where this approach really shines: this is the same basic technique (though more complicated) used to coat aluminum on mylar to make those shiny metallic balloons. Thin film deposition is very well understood, there's lots of off-the-shelf equipment, and it's suitable for making vast rolls of material in a continuous process.

Right now, there are some thin film solar cells on the market, but the price is still an order of magnitude above traditional solar cells. This may partly be a function of low manufacturing volume (thin films really shine when made in truly mass quantities), and it may also be a function of immature technology. Some of the articles I've read suggest that manufacturing yields are still low, and durability under real-world conditions is still a big problem.

Photochemical Cells

Photochemical cells don't directly produce electricity, but instead use solar energy to produce a chemical reaction which can yield energy. The most research seems to be into photochemical cells which generate hydrogen gas from water (the hydrogen can then be consumed in a fuel cell to produce electricity, stored for later use, or used as a somewhat impractical fuel for vehicles). A catalyst submerged in water absorbs sunlight, and uses the energy to split water molecules yielding hydrogen and oxygen.

In the lab, photochemical cells have exceeded 10% efficiency, which makes them very promising. The problem is that the materials used to promote the chemical reaction tend to corrode in water. You get 10% conversion efficiency, but the electrodes only last a month or so: not very practical for the real world, especially if the catalysts are made of exotic materials.

On the other hand, if the durability issue can be fixed, photochemical cells have considerable potential. In addition, since the hydrogen gas can be stored for later use, it eliminates the problem of only having power available on sunny days.

Dye-Based Solar Panels

Another promising approach is the use of a photosensitive dye to absorb light and convert it into usable energy. Solar cells need to do two things: generate free electrons, and enforce a charge separation so that the electrons accumulate and can be used to do work (instead of just recombining right away). In a traditional solar cell, the same material serves both functions, so you need something which is efficient at both absorbing light and separating the charges.

A dye-based solar cell uses two materials: the dye (which absorbs light and generates electrons) embedded in a semiconductor matrix which creates the charge separation. This allows a much wider variety of materials, and much less expensive materials like Titanium Dioxide. Manufacturing can also be less expensive, since it can use techniques developed for processes like printing.

This approach also seems to have an advantage in low-light conditions: where silicon has a threshold intensity to produce power, the dye-based solar cells will keep producing a small amount of power even in low light. This could yield a significant boost in total annual power production, especially in cloudy climates and high latitudes.

According to the articles I've read, the first manufacturer of dye-based solar cells, G24 Innovations, is just now beginning production. I wasn't able to find any actual products for sale, so I don't know how the prices compare to traditional solar cells.

Other Approaches

Anyone with a cheap pocket calculator can do the calculations I've done which show that a substantial (though not impossible) improvement in solar cell technology will truly change the energy game--and not incidentally make a ton of money for the inventor.

Pretty much any mechanism which can turn a photon into a free electron is being pursued somewhere as a new solar technology. I'm reminded somewhat of the story of the invention of the light bulb: the carbon arc lamp proved the usefulness of a small clean-burning electric lamp, and inventors were working on the problem for a long time before Thomas Edison finally managed to make it practical.

Similarly, today's expensive silicon-based solar cells prove that there's a vast market for a cheap way to convert sunlight into electricity. If it can be done, someone will find a way, and probably within a few decades.