If we could devote roughly two percent of the U.S.’s land area to generating electricity by solar means, that would meet the country’s energy needs, says MIT’s Emanuel Sachs. This doesn’t mean just its electricity needs, he notes: “I’m talking about all our energy needs.”

Sachs, a mechanical engineering professor, says the notion isn’t as implausible as it may sound — especially considering we already devote nearly that much land to roads and highways. On the other hand, it would require transitioning huge segments of the country’s energy system, including cars and trucks, to electricity. It would also require massive progress in our ability to store power.

“You could send 10 to 15 percent of the amount of electricity the country uses now directly out on the grid,” he notes, “but the rest you’d have to store.” Still, he argues that focusing on solar’s potential tells an important story.

One barrier to a dramatic expansion of solar is that photovoltaic power — electricity produced from light — runs roughly four times the cost of electricity from coal-fired power plants. This hurdle, though, may be surmountable. In fact, Sachs has spent much of his career trying to cut the costs of making the devices that make the electricity, because that’s where the biggest savings are.

One key business part of today’s solar panels, called a cell, is made of refined silicon — the same stuff as electronic chips. And while alternative materials are under development in other labs, Sachs says silicon has some advantages. For one, the industry knows a lot about it. For another, it’s Earth’s second most abundant element.

The latter point could be critical if solar were ever to become the major energy source. But of course, the cost-effectiveness of photovoltaic power has to go way up first.

How could that happen? Boosting the performance of solar products is only part of the answer, says Sachs. “Today’s best commercial systems are about 18 percent efficient,” he notes, “and the theoretical upper limit for crystalline silicon is about 29 percent.” The other part, of course, is conceptual breakthroughs in the manufacture of silicon-based systems.

Sachs himself has made a notable head start on that goal. The silicon in solar panels has traditionally been created in the form of one-at-a-time wafers — very thin slices pared item by item from a block of expensive refined silicon. A portion of each block, though, ends up as useless dust. “Most manufacturers today effectively throw away half their refined silicon,” he notes.

Sachs’ solution? Create the silicon industry’s equivalent of one of those soapy-water wand-and-dish combinations found in kids’ stores. But instead of creating filmy sheets of soapsuds, the approach involves dipping pairs of high-temperature filaments into vats of molten silicon. The result is a continuous, three-inch wide ribbon of silicon that is then laser-cut into wafers.

The system’s already in use at Massachusetts-based Evergreen Solar. But this advance, says Sachs, reflects just a hint of what’s achievable with a sustained drive to cut the expenses involved in preparing and forming the silicon.

Would such an effort really be able to trim costs by a factor of four? “If you look at the fundamental structure of this issue — how much does the silicon cost, how much do you need — the answer is clearly ‘yes,’” says Sachs.