Skip to Content
MIT News magazine

Bright Days for Solar

New investment fuels promising energy research at MIT.
August 19, 2008

When Tonio Buonassisi finished his PhD at the University of California, Berkeley, in 2006, he wanted out of academia. He had a mission: to cover the world’s roofs with solar panels that would provide clean, ­carbon-­free electricity. A cash-strapped university research program wasn’t the place to make that happen. But research dollars were ­plentiful in the solar industry, which was doubling in size every couple of years. Having “lived through times of famine” as a grad student, Buonassisi says, he decided against taking a postdoc appointment and went to work at a manufacturer of solar panels based in Marlborough, MA.

Sunny Side of the Street: Tonio Buonassisi is optimistic that innovation in the lab can make solar energy affordable.

Less than three years later, Buonassisi is back in academia as an assistant professor of mechanical engineering at MIT, which has seen a new surge in solar research propelled by the MIT Energy Initiative and a growing student energy club. For decades, the only solar research at MIT had been carried out intermittently by a small cadre that Buonassisi describes as “a few lonely torch-­bearers wandering around in the caves.” A “dead period” descended on the worldwide solar-research community after both oil prices and interest in alternative energy plummeted in the mid-1980s, he says. But now the world is clamoring for affordable clean energy. And thanks in large part to the establishment of the MIT-­Fraunhofer Center for Sustainable Energy Systems (where Buonassisi is scientific director) and the Solar Revolution Project, tens of millions of dollars earmarked for solar research have started to pour into the Institute from both government and private sources. About 20 MIT researchers have turned their attention to solar energy so far.

The potential of solar power is enormous. In just over one hour, enough energy from the sun reaches the earth to meet the world’s needs for an entire year. But as of yet, that potential is far from being realized. Solar panels, which convert the energy in sunlight into electricity, are so expensive and inefficient that they now produce just a fraction of 1 percent of the world’s power.

That has to change if the world is to hit widely accepted targets for cutting carbon dioxide emissions while meeting growing energy demand. The targets call for at least 10 million megawatts of ­carbon-­free generating capacity by 2050, most of which will probably need to come from solar, given the limits of other clean energy technologies. Solar power now provides only about 10,000 megawatts worldwide. To reach 10 million megawatts, the solar industry will need to continue its rapid growth for the next 20 to 30 years.

Some of the research at MIT could have an almost immediate impact on the industry. One example is Buonassisi’s effort to improve the price and efficiency of solar cells, the ­semiconductor-based components–most often made of silicon–that convert light into electricity in a solar panel. (Each panel is made of many cells packaged together with electronics and a frame.) Other projects may take longer to complete but could yield big results. One group of researchers is pursuing a way to concentrate sunlight using luminescent dyes developed for next-generation flat-screen displays; such technology could greatly reduce the amount of expensive silicon needed to generate electricity from the sun. Another professor is creating catalysts that imitate photosynthesis, using sunlight to make chemical fuels. That could solve one of the basic problems with solar energy today: there’s no good way to store it.

Cheap and Dirty Silicon
In spite of its predominance in the industry, in many ways silicon is a lousy material for solar cells. Because it doesn’t absorb light as well as some other semiconductors, a relatively thick slab of silicon is needed to generate useful amounts of electricity. But using a thick material is problematic. When a silicon solar cell absorbs light, the light’s energy frees electrons to move through the material. To generate electricity, the electrons have to escape the material and reach an outside circuit. If the silicon is thick, the electrons have far to go to escape. Along the way, they can be trapped by defects and impurities in the material. So the silicon must be made as defect free and pure as possible–an expensive proposition.

Yet if solar energy is ever going to be widespread, then silicon cells are a good choice, because silicon is extremely abundant–second only to oxygen in the earth’s crust. The current leading alternatives to silicon-based solar cells use rare elements, significantly limiting their potential for large-scale use.

So Buonassisi and fellow mechanical-engineering professor Emanuel Sachs are developing techniques to overcome silicon’s shortcomings. Sachs recently started 1366 Technologies, based in Lexington, MA, which will commercialize advances from his lab that make silicon solar cells more efficient. For example, much of the light that enters a conventional silicon solar cell reflects back out again. Sachs has found a way to manufacture solar cells so that more of the light that enters them bounces around until it’s absorbed and converted into electricity.

Meanwhile, Buonassisi is figuring out how to keep contaminants and defects in silicon from trapping electrons. Today solar-cell manu­facturers use the same expensive high-grade silicon that goes into computer chips. But while the chip industry spends about $1,000 per gram on clean manufacturing processes to keep its silicon pure, solar-cell companies are limited to less than a dollar per gram, Buonassisi says; otherwise their products can’t compete with conventional sources of electricity. And a dollar a gram isn’t enough to keep silicon clean. At one stage in the manufacture of a solar cell, for example, the silicon is heated in a crucible to about 1,400 ºC, before being cooled again to make crystalline ingots. But iron from the crucibles and stainless-steel furnace parts dissolves in the silicon. “You take your ultrapure silicon,” Buonassisi says, “and then you dump it into a very dirty production environment.”

Buonassisi sees such impurities as unavoidable but manageable. He found that if the superheated silicon is cooled to about 500 ºC and held there, the dissolved iron will precipitate out, move through the silicon, and form clusters of iron silicide. Gathering the iron atoms together in this way makes it less likely that electrons will run into them, so more electrons can escape to produce an electric current. Holding the silicon at 500º for about 30 minutes can improve the efficiency of a solar cell by 3 to 7 percent. A solar-cell manufacturer could sell these more efficient cells at higher prices, generating enough profit in two years to build a new manufacturing plant. With more plants running, manufacturers could significantly speed up the production of solar cells.

By focusing on ways to minimize the impact of impurities, Buonassisi’s research could also allow manufacturers to make solar cells out of cheaper, dirtier silicon. And cheaper raw materials combined with greater production capacity would put solar power in a better position to become a major source of electricity.

Colorful Concentration
Vladimir Bulovic, an associate professor in the department of electrical engineering and computer science, inserts a copy of Pirates of the Caribbean into a Blu-ray disc player, and the screen in his office–an elegant device just a few millimeters thick–comes to life. The remarkably vivid and crisp colors look good from any angle. The images appear touchably real.

An expensive Sony showpiece, the next-generation display uses organic light-emitting diodes as pixels. Although Bulovic’s research helped make the display possible, he’s not just showing off his handiwork. The organic molecules at the heart of the device could be the key to making solar power as cheap as power from coal.

In traditional solar panels, a semiconductor is spread out over the entire surface to absorb sunlight. But for decades, engineers have known that curved mirrors or lenses can gather sunlight from a large area and direct it to a small solar cell, enabling a fraction as much semiconductor material to absorb the same amount of light. What’s more, solar cells perform better under concentrated light, converting a higher percentage of photons into electricity.

This approach has its drawbacks, however. Bulky mirrors and lenses can’t be installed on rooftops and are easily damaged by wind. They also require mechanical tracking systems that keep the sunlight focused on the small solar cell throughout the day. These systems add cost and tend to break down. In the 1970s, scientists proposed a simpler scheme that would use light-emitting organic molecules to concentrate sunlight. Such materials–like the ones in Bulovic’s display–are now available for the first time.

MIT electrical-engineering professor Marc Baldo, with whom Bulovic is collaborating, has championed a way of using the organic molecules in solar technology. It goes like this: Coat a glass sheet with the molecules, then expose the glass to sunlight. The mole­cules will absorb the light and reëmit it at another wavelength. Because light moves at different speeds in glass and in the surrounding air, the reëmitted light reflects back into the glass at the boundary between the two. About 80 percent of the light the mole­cules emit will bounce around inside the glass until it reaches the edges of the sheet. (The same phenomenon allows light to travel through fiber-optic cables.) As a result, the sheet of glass gathers light and concentrates it at the edges. Solar cells just a couple of millimeters wide (in most solar panels, they’re several centimeters across) can be laminated onto the edge of the glass to absorb the light and generate electricity. A sheet of plastic with the organic molecules embedded in it can concentrate light in the same way.

Of course, luminescent organic molecules–better known as luminescent dyes–have been around for a long time. But they haven’t worked for concentrating sunlight because they tend to fade after prolonged exposure and to reabsorb much of the light they emit, preventing it from reaching the edges of a sheet. Recently, however, researchers have made luminescent display dyes durable enough to survive for years or even decades in direct sunlight, making them viable for use in solar applications. What’s more, Baldo has demonstrated a way to use dyes that don’t absorb the colors they emit, which allows more light to reach the edges.

Unlike existing solar concentrators that use lenses and mirrors, sheets of luminescent glass or plastic wouldn’t have to be pointed directly at the sun to absorb sunlight, eliminating the need for tracking. The sheets would also be lighter and easier to install than a comparable parabolic mirror or lens. The technology could even be built into homes, where windows “painted” with the transparent dyes could be edged with strips of solar cells.

Storing Light
Even if solar panels become extremely cheap and efficient, they still have a problem: they reach their peak power output for only a few hours a day and don’t work at all at night. As long as solar energy provides only a small fraction of the world’s electricity, other sources–often fossil fuels–can be relied upon to supply power after dark. But as solar grows, finding a way to store the energy it produces will become essential.

Researchers are investigating many ways to do this. At MIT, for example, Donald Sadoway, a professor of materials chemistry, is working on new kinds of high-capacity batteries. And ­Daniel ­Nocera, the Henry Dreyfus Professor of Energy, is developing inexpensive ways to use sunlight to split water into hydrogen and oxygen, which can be recombined in a fuel cell to generate electricity.

The standard process for splitting water, called electrolysis, is inefficient and expensive. That’s why almost all the hydrogen produced today comes from fossil fuels, not water. But Nocera is taking a cue from nature, looking to photosynthesis for ideas about cheaper, more efficient techniques for splitting water.

In photosynthesis, plants use sunlight’s energy to generate electrons and their positively charged counterparts, holes. An enzyme called the oxygen-evolving complex uses the holes to produce oxygen gas and hydrogen ions from water. Another enzyme combines the hydrogen ions with electrons to make hydrogen.

Until a couple of years ago, researchers didn’t clearly understand the oxygen-evolving complex or how it works. When its structure was discovered, Nocera’s lab flew into a frenzy of activity aimed at developing an artificial version of the enzyme that would be cheap and easy to make on a large scale. This spring, Nocera announced success: a new catalyst made from readily available cobalt can produce oxygen from water using sunlight.

More work remains, since converting the hydrogen ions into hydrogen gas is an extra step requiring a costly platinum catalyst. But Nocera says the discovery of a catalyst for making oxygen has supplied the missing link in the quest to mimic the action of the leaf.

“You can make fuels from a glass of water with sunlight now very cheaply,” he said at a recent symposium. Nocera predicts that in 10 years, it could be practical for homeowners to install systems that use photovoltaics to generate power during the day and split water to make fuel at night.

Gaining momentum
On May 3, 1978, President Jimmy Carter announced a new push to make solar power cheap enough to compete with electricity generated from fossil fuels. He said that goal could be reached by 1990. But within just a few years, investment in solar panels had waned, and solar research had fallen into Buonassisi’s “dead period.”

The researchers now exploring solar energy at MIT are picking up where the surge of research in the late 1970s and early ’80s left off. Like Vladimir Bulovic’s work on luminescent concentrators, Buonassisi’s method for removing defects from crystalline silicon has its roots in the ’70s: it was suggested by the research of Ali Argon, now an MIT professor emeritus who has advised ­Buonassisi. And Emanuel Sachs is one of MIT’s solar torch-­bearers: his work in the early 1980s led to the founding of Evergreen Solar, the company that lured Buonassisi away from academia.

The difference is that this time around, the solar industry is orders of magnitude larger, Buonassisi says. “If you feel that scale and put your finger to that pulse, you start to have respect for the momentum that this industry has behind it now,” he says. “It didn’t have anything close to that in the 1970s.”

So as the work at MIT expands, the idea of making solar power competitive with conventional sources of electricity in much of the United States looks increasingly feasible, even on Carter’s ambitious 12-year time line. And that has researchers increasingly excited. “We have a $50 billion industry behind us,” Buonassisi says. “There’s large potential for anything we do in the laboratory. A little bit of innovation can have an enormous impact.”

Keep Reading

Most Popular

10 Breakthrough Technologies 2024

Every year, we look for promising technologies poised to have a real impact on the world. Here are the advances that we think matter most right now.

Scientists are finding signals of long covid in blood. They could lead to new treatments.

Faults in a certain part of the immune system might be at the root of some long covid cases, new research suggests.

AI for everything: 10 Breakthrough Technologies 2024

Generative AI tools like ChatGPT reached mass adoption in record time, and reset the course of an entire industry.

What’s next for AI in 2024

Our writers look at the four hot trends to watch out for this year

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.