When it comes to the photon-collecting element of photovoltaic systems, as with so many other areas of technology, more is generally better. But if two MIT researchers succeed in their current quest, less could soon have its, ahem, time in the sun.
The researchers are Steven Johnson and Peter Bermel. And underlying the pair’s current work is a vexing performance gap between so-called thin-film photovoltaic materials — most either of silicon or of selected inorganic compounds, like cadmium telluride — and conventional solar-cell designs.
Solar cells whose semiconductors are, speaking strictly in comparative terms, pretty thick, are pretty good at absorbing sunlight across and even outside the visible spectrum. “Silicon that’s 200 microns (about 1/70th of an inch) thick will absorb just about everything that’s within the material’s absorptive capacity,” notes Johnson, an MIT assistant professor of applied mathematics. This includes the entire visible spectrum, from violet to red, as well as some infrared light.
But at such dimensions, refined silicon’s high price tag is a big problem. Bermel, an MIT Research Laboratory of Electronics “postdoc” who works with Johnson, notes that “for some manufacturers, silicon is 25 percent of the cost of their systems.”
Cost issues are driving selected manufacturers to thin-film products: materials one-hundredth or less as thick as conventional photon-absorbing semiconductors.
Thin-film photovoltaic semiconductors are made using processes like chemical vapor deposition, which are comparatively cheap.
But if you use thin films, you’re in effect writing off a large share of the sun’s light. “Once you get to the red end of the spectrum, thin films won’t absorb,” notes Johnson.
What inspired Johnson’s and Bermel’s project was the fact that, generally speaking, the farther light travels within a medium, the more gets absorbed. So the pair’s strategy for capturing photons at the red end of the spectrum is to bounce the light that penetrates the thin film back into the semiconductor, but at an angle such that a lot of the returning light travels the long axis of the thin-film material.
“You basically want to redirect the light so it nearly goes sideways,” explains Johnson. And the researchers had the technology they thought would do the trick.
The approach reflects advances in photonics, a field whose researchers have created astonishing techniques for manipulating light: “You can stop light, you can trap it, and you can create all sorts of other effects,” notes Johnson.
To get the specific type of light behavior they sought, the MIT researchers are creating a photonic crystal coating for the bottom surface of thin-film silicon. The coating, itself hyperthin, is made up of a 3-D crystal lattice. Several such lattices have been proposed over the years, and one of those the MIT researchers are exploring happens to be reminiscent of certain produce displays.
Think of millions of microscopic glass spheres packed together like oranges on a supermarket case. You fill the spaces between the spheres with a material of your choice. Then, you eliminate the orange equivalents, i.e., the spheres.
The solid part of the lattice could be any of a number of materials, but silicon itself has strong appeal. “The system’s effectiveness depends not on the materials themselves but on the interfaces between them,” explains Johnson, “so you want two components — for example, air and silicon — that are as different optically from one another as possible.”
It has taken lots of work to hit on coatings that seem to do what the researchers want: reflect a modest amount of light mirror fashion; and, bend the remaining light so it travels the long way in thin films.
How the photonic crystal traps and reorients the photons is a complex story. The key point, though, is that because the ways in which the light gets bent largely depend on the specific size and arrangement of the structure, the crystal has to be manufactured to an almost unimaginably demanding set of specifications.
Experiments so far have been promising. “A thin-film silicon cell with no enhancements is about five percent efficient,” notes Bermel. “The most advanced thin-film systems are eight to nine percent, and we project that ours will be about 13 percent efficient.”
This says the approach has marketplace potential, and in fact Bermel’s involved with a startup. But it’s unlikely to be a short journey to an actual set of products, especially given the minuscule dimensions of the 3-D crystal’s internal architecture. “It’s amazing that you can even think about manufacturing something on the length scales we’re dealing with,” notes Johnson.
Still, the researchers like the idea of moving a product from lab to marketplace. Johnson’s been involved in that transition before: a photonic crystal-based endoscopic system he worked on is now helping patients with conditions like esophageal cancer.
“If you have a problem whose solutions can improve people’s lives,” he notes, “it’s a big motivating factor.”