All of the Earth’s energy is “solar” in the sense that its ultimate origin lies in the sun. However, with petroleum resources dwindling worldwide, people can’t wait several hundred million years for the processes of natural photosynthesis to replenish the world’s stocks of fossil fuels.
Instead, researchers such as Vikram Dalal, Thomas M. Whitney Professor of Electrical and Computer Engineering, and Rana Biswas, an ECpE adjunct associate professor and affiliate of the Institute for Physical Research and Technology, are seeking ways to access the sun’s energy more directly through cutting-edge photovoltaic technologies.
With a significant grant pending from the Iowa Power Fund, the researchers and their collaborators seek to develop thin-film solar cells that incorporate amorphous silicon and nanocrystalline silicon into a supercell that can achieve levels of photonic conversion typically associated with standard crystalline silicon cells. By exploiting the novel technology of photonic band gaps and plasmonics, the team hopes to make solar technology far more efficient at a fraction of the cost.
That’s a near-term goal. In addition, their work under the Iowa Power Fund will look 15 years out to more exotic approaches such as organic semiconductors that essentially reverse the mechanism of the organic light-emitting diodes (LEDs) now coming to market to instead convert sunlight into electricity. A hybrid form of reverse-engineered LED, says Dalal, could take advantage of the ability of organic semiconductors to absorb light, while compensating for their less robust ability to transmit electrons.
Ancient technology, modern challenge
Although heating buildings using solar energy has been practiced since ancient times, Dalal notes, not even the 1970s oil embargo was enough to drive western consumers to adopt solar technologies. While cost-efficient in sunnier climes, he says, natural gas at $1 per 1,000 cubic feet or oil at $3.70 per barrel took the shine off comparatively expensive solar technologies in the Midwest and Northeast.
Today, with gasoline as high as $4 and natural gas at $8 to $10 or more per 1,000 cubic feet, solar technologies are looking far more attractive to consumers.
“There is enough solar energy—even in Iowa—to be able to heat your home comfortably,” Dalal observes, “with gas providing only an assist during bitter cold periods or when the sun isn’t out.”
Modern solar technologies take two basic forms, according to Dalal: Using a parabolic trough, you can concentrate the sun’s rays to boil water to run a steam turbine, a technique employed by large solar “farms” that generate electricity for the grid. Or, you can directly convert photons of light into electrons using a semiconducting material such as silicon—in short, a simple solar cell of the type developed in the 1950s for use on the Vanguard satellite.
“The problem was that the cost of solar cells designed for satellites was around $100,000 a kilowatt (KW) as late as the 1970s,” says Dalal. “Coal-fired power plants were around $1,000, so you had 100 times the cost for solar. Today, solar panels using semiconductors to convert solar energy directly into electricity have come down from $100,000 a KW to $3,000.”
In a climate such as California or Arizona, a $3,000/KW solar panel can translate into as little as fifteen cents per kilowatt hour (KWh), amortized over the life of the panel. That could make a typical installation competitive with the current average costs per KWh from other sources, which are likely only to rise in the future—Dalal notes that a new coal-fired plant recently approved for Marshalltown, Iowa, has been estimated at $2,600/KW.
Yet while conventional solar cells are approaching economic viability on an amortized basis, the up-front costs are still sufficiently daunting to discourage most homeowners and developers, adding tens of thousands of dollars to the price of a new home. In order to make such an investment more attractive, then, a less-expensive alternative to conventional crystalline silicon semiconductors must be found.
Order at the nanoscale
If, Dalal asserts, the cost of solar cells can be reduced from the current average of $3,000/KW to $1,000, plus an average installation cost of $1 per watt, not only would up-front costs be a fraction of those for photovoltaic cells from crystalline silicon, the immediate cost of energy production would be less than that of electricity from coal-fired plants—as little as 6 or 7 cents per KWh.
Based on lattices of amorphous and nanocrystalline silicon, the thin-film solar panel being developed by Dalal and his MRC team initially will not have the efficiency of standard crystalline silicon cells—that’s perhaps 15 years out, Dalal acknowledges. But, he feels, the much lower cost of production will buy time to refine the technology to the point where it has similar efficiency but at a much lower cost.
The challenges are considerable. Amorphous silicon is what Dalal calls a “disordered semiconductor” that, while highly efficient for absorbing photons, suffers from discontinuity at the nanometer scale, posing challenges for the transport of electrons. While this can be remedied somewhat by introducing hydrogen into the substance, hydrogenation makes the silicon more vulnerable to degradation from sunlight.
By contrast, nanocrystalline silicon (NCS) is a highly ordered and stable environment in which electrons behave as if they were in crystalline silicon. However, because electrons exhibit confusion at the grain boundary in NCS, appropriate passivation techniques using very thin films of amorphous silicon between the grains are needed to improve transport of electrons from one grain to the next.
Dalal’s MRC team had already introduced certain modifications to these materials in advance of the current power fund project. In 2005, for instance, Dalal employed a chemical annealing process to reduce the degradation from sunlight of single-junction amorphous silicon in the first year of application from the normal 15%–20% to about 4%.
“We want to implement this superior sun-proofing technology into our amorphous silicon cells so that the combination of amorphous silicon and nanocrystalline silicon will degrade only 2%,” says Dalal. “Who cares about 2%? It’s insignificant. Even crystalline silicon degrades by 2%.”
Tailoring the band gaps
Degradation issues aside, though, unless the two forms of silicon together can function as efficiently as cells based on a normal crystalline silicon architecture in capturing and converting sunlight into energy, whatever savings they might realize in production would be lost in application, making their thin-film technology uncompetitive. As Biswas notes, meeting the challenge requires reconciling two competing imperatives.
“On the one hand,” Biswas says, “the cells cannot be very thick, because once electrons are created inside the cell by sunlight, they have to travel to the electrode. And the diffusion rate of electrons for amorphous silicon is about 3/10 of a micron. Once you go beyond that, the electrons won’t get captured.
“But as you make the cells thinner,” he continues, “you lose the ability to capture enough light. So the hurdle is to actually redesign the cell so that the photons have enough path length, enough distance to travel within the cell. That’s what we’re doing in this project as well.”
One technique for circumventing the problem is through the exploitation of photonic band gaps, which either allow or disallow the transmission of photons from different regions of the solar spectrum. Most of the spectrum’s intensity is concentrated in the visible region, Biswas observes, so it’s especially important to capture the energy there.
“The rough idea is that you have layers of different band gaps that are absorbing different regions of the solar spectrum,” Biswas says. “Typically, the high band gaps would absorb the blue and green parts of the spectrum and the low would concentrate more on the red.”
Next, according to Dalal, if you have a regular structure of refractive indices in two dimensions, you can force photons to be reflected back into the material for 100% efficiency; that is, they will not migrate back into the photonic band gap, where they would be unavailable for conversion into electrons. Using this method, photons can achieve longer optical paths, effectively making a two-micron film behave as if it were 20 microns in thickness.
“That way you can increase photon absorption significantly without increasing the amount of material you use,” Dalal remarks. “So you use 100 times less silicon than a crystalline wafer, yet achieve nearly the same efficiency.”
A $100 billion industry
Further efficiencies, Dalal notes, may be achieved through improving and enlarging the grain structure of NCS, all of which will culminate in the development of novel device structures that increase efficiency by building “super-lattices” of one material followed by another—for example, 50 cycles of alternating NCS and amorphous silicon. It’s a structure, he suggests, that may in the lab result in a conversion efficiency of 20%—fully the equal of crystalline cells—within the next five years.
These are ambitious goals for a fledgling technology, Dalal admits. But the stakes are huge, and Iowa Power Fund’s investment in research devoted to applications aimed at returns in the near term is a reflection of the state’s economic development priorities. Likewise, the project’s partnership with local industries such as PowerFilm of Ames, Iowa, underscores the three-way alliance of government, industry, and academic research.
“After all, the reason that Boston and northern California are virtually immune from recession is the constant spinoff of industries from universities such as Stanford, Berkeley, MIT, and Harvard,” Dalal stresses.
“Solar is going to be a $100 billion industry,” he adds. “If Iowa gets 5% of that, that’s $5 billion. But in order to do that, we have to nurture research at Iowa’s universities.
Vikram Dalal and Rana Biswas are just two collaborators on the anticipated Iowa Power Fund project for new solar technologies, now in the final stages of negotiation. Others working on the project include Jaeyoun Kim and Sumit Chaudhary of the Department of Electrical and Computer Engineering, Malika Jeffries-El of the Department of Chemistry, Joseph Shinar of the Department of Physics, and Ruth Shinar of the Microelectronics Research Center.