Gregoire's CleanTech Blog

Novel Batteries Get a Boost

General Motors has announced it will work directly with A123 Systems, a start-up based in Watertown, MA, to develop batteries that are fine-tuned for the Chevrolet Volt, an electric vehicle scheduled for production in 2010 or 2011. The new agreement between the companies is designed to speed the vehicle to market.

The co-development deal is the latest in a series of announcements by the automaker indicating it is moving forward with plans for the Volt, an extremely fuel efficient hybrid. The Volt, first announced in January, will use batteries that can be recharged from a standard electrical outlet and is expected to get 40 miles on a single charge, eliminating trips to the gas station for average daily driving. While GM has not ruled out using batteries from another supplier, the agreement is a vote of confidence that A123's technology can meet the requirements for the vehicle, says Denise Gray, GM's director of hybrid energy storage devices.

A123 uses a new lithium-ion chemistry that allows its batteries to be much lighter and more compact than the nickel metal hydride batteries in existing hybrids today, and safer than the conventional lithium ion batteries found in consumer electronics. In June GM announced that it is working with the South Korean company LG Chem, and its subsidiary Compact Power, based in Troy, MI, to make both battery packs and the individual cells inside them. They also signed an agreement with an LG Chem competitor, the Frankfurt, Germany-based Continental Automotive Systems, to develop battery packs. Continental had planned to use A123 as a subcontractor to supply the batteries for these packs. The new agreement puts A123 in direct contact with GM on the Volt project.

At the end of 2005, A123 announced its first product, a cylindrical battery for power tools. For the Volt, A123 is shifting to a much bigger and flat, rather than cylindrical, battery cell that takes up significantly less space. Hundreds of these would be combined in a large pack, with control electronics and a cooling system, to power the Volt.

A123 employees have begun testing the new cell. In one test, a prototype cell, a silvery slab about the size of three decks of cards placed side by side, was crushed from the side to demonstrate its abuse tolerance. Unlike conventional lithium-ion batteries, which use a cobalt-oxide material that can overheat and cause laptops and cell phones to burst into flame, A123's batteries use a much more stable, and potentially cheaper, iron phosphate material. The crushed test battery showed no signs of overheating.

Earlier this year GM had picked A123 Systems as a potential supplier of lithium-ion batteries for a new version of its Saturn Vue hybrid that can be recharged by plugging it in, giving it a 10 mile all-electric range. It could be on the roads as early as 2009.

But the battery requirements for the Volt are much more challenging, requiring the different cell design. Whereas in the Vue power from the battery will be supplemented with power from a gasoline engine, the Volt will rely on batteries full-time. For the first 40 miles of driving, the battery will deliver power from being plugged in overnight. After that, the battery will continue to power the vehicle as it is recharged by an on-board gasoline-powered generator.

The new batteries must pack more energy into a smaller space, the reason for the flat design. But they also need to be fine-tuned for the power demands of the Volt. To do this, engineers can subtly alter the electrode chemistry, thickness, the type of electrolyte, and other aspects of the battery cell design. The collaboration with GM will give A123 direct access to more lab space and testing equipment that could allow A123 to more quickly sort through many subtle variations on the cell design, says David Vieau, A123's CEO. "You never get enough test capacity for all of the different variations you'd like to be able to consider," says Vieau. "We hope this [agreement] will accelerate the process."

Gray says GM hopes to have the batteries in packs within the next couple of months and to have these packs in a generic vehicle platform for testing by the end of the year. The technology could be in a prototype Volt sometime next spring, she says. GM's goal of a production model by 2010 or 2011, Gray says, is an "aggressive target." But the tests they've done so far since announcing the project in January have been encouraging. "We've got no information back that says uh-oh we can't do this," Gray says.

The Volt is going to be about the size of a Chevrolet Cobalt, which currently costs between $14,000 and $20,000 depending on the configuration. Rob Peterson, a GM spokesperson, says GM won't "price themselves out of that class" of vehicles with the Volt. He adds that if GM intended the vehicle to have a hefty premium, it would have made the vehicle a Cadillac. "It was a conscious decision to make it a Chevy," he says.

The agreement with GM does not mean that A123 will be locked out of doing work with other automakers. The specific cell the two companies develop together will be co-owned--so both companies will have to negotiate for how it might be used, Gray says. That could include putting the pack in cars by other companies, Gray says, since "it's in everyone's interest to achieve economies of scale to drive down battery prices." A123 still retains control over its novel battery chemistry, however. Indeed A123 has had discussions with other automakers and its batteries are set to be used in hybrid buses starting in 2008.

Silicon Nanocrystals for Superefficient Solar Cells

A typical solar cell generates only one electron per photon of incoming sunlight. Some exotic materials are thought to produce multiple electrons per photon, but for the first time, the same effect has been seen in silicon. Researchers at the National Renewable Energy Laboratory (NREL), in Golden, CO, showed that silicon nanocrystals can produce two or three electrons per photon of high-energy sunlight. The effect, they say, could lead to a new type of solar cell that is both cheap and more than twice as efficient as today's typical photovoltaics.

As in earlier work with other materials, the extra electrons come from photons of blue and ultraviolet light, which have much more energy than those from the rest of the solar spectrum, especially red and infrared light. In most solar cells, the extra energy in blue and ultraviolet light is wasted as heat. But the small size of nanoscale crystals, also called quantum dots, leads to novel quantum-mechanical effects that convert this energy into electrons instead.

By generating multiple electrons from high-energy photons, solar cells made of silicon nanocrystals could theoretically convert more than 40 percent of the energy in light into electrical power, says Arthur Nozik, a senior research fellow at NREL. In contrast, today's flat rooftop solar panels are at best just over 20 percent efficient and are theoretically limited to about 30 percent efficiency. Concentrating sunlight with mirrors or lenses could raise that figure to about 40 percent, but the same approach could boost the efficiency of a silicon-nanocrystal solar cell to well over 60 percent, Nozik says.

What's more, solar cells made of silicon nanocrystals could prove to be cheap, giving them a significant advantage over other approaches to high-efficiency solar cells. For example, advanced "multijunction" cells have shown efficiencies of more than 40 percent. But these require complicated manufacturing processes that combine expensive semiconductors optimized for different parts of the solar spectrum. Silicon nanocrystals, in contrast, are relatively easy to make, even compared with the material in conventional solar cells, the best of which are made of very large, single crystals of silicon.

Silicon nanocrystals also have marked advantages over the other nanocrystal materials that have shown the multielectron effect. Some of these materials contain toxic elements such as lead or cadmium, and others rely on elements such as indium that are in limited supply. But silicon is both safe and abundant. It's also well studied, says Christiana Honsberg, professor of electrical and computer engineering at the University of Delaware, so engineers know how to work with it to make solar cells. Indeed, for many of the same reasons, silicon is by far the most common material in solar cells today, and it's attractive as the basis for broader deployment of photovoltaics in the future.

Before the NREL work, researchers had believed that silicon crystals small enough to produce the multielectron effect would be impractical as a photovoltaic material. At the nanoscale, the optical properties of silicon change so that it converts less light from the red end of the spectrum into electrons. As a result, any gains from more efficiently converting blue and ultraviolet light would be offset. Nozik and his colleagues found that the nanocrystals did not have to be as small as was previously thought, skirting this problem.

To be sure, the NREL work is only a first step. Making solar cells that take advantage of multielectron generation is a challenge. That's because the extra electrons are very short-lived, making it difficult to extract them from the nanocrystals to generate an electrical current. Indeed, this has proved so difficult that evidence of the effect has come from indirect methods such as spectroscopy rather than from current generated by a solar cell. The use of the indirect measures has led some prominent experts to question whether the extra electrons are actually being produced, although Nozik says that the effect has been confirmed using multiple techniques. Nozik and his colleagues are now working to make solar cells out of silicon nanocrystals--they're exploring a number of novel designs--and he says they've recently made direct measurements indicating that their cells are releasing multiple electrons per photon absorbed. (Their results have yet to be published.)

Honsberg is cautiously optimistic, calling the finding of the multiple-electron effect in silicon nanocrystals a breakthrough, but only "one breakthrough out of maybe three or four" needed to produce cheap, superefficient solar cells.