Copper power: This prototype solar cell uses a copper-based material and has achieved record efficiencies for a cell of its kind.
IBM Research
IBM researchers have greatly increased the performance of a novel thin film solar cell.
Researchers at IBM have increased the efficiency of a novel type of solar cell made largely from cheap and abundant materials by over 40 percent. According to an article published this week in the journal Advanced Materials, the new efficiency is 9.6 percent, up from the previous record of 6.7 percent for this type of solar cell, and near the level needed for commercial solar panels. The IBM solar cells also have the advantage of being made with an inexpensive ink-based process.
The new solar cells convert light into electricity using a semiconductor material made of copper, zinc, tin, and sulfur--all abundant elements--as well as the relatively rare element selenium (CZTS). Reaching near-commercial efficiency levels is a "breakthrough for this technology," says Matthew Beard, a senior scientist at the National Renewable Energy Laboratory, who was not involved with the work.
The IBM solar cells could be an alternative to existing "thin film" solar cells. Thin film solar cells use materials that are particularly good at absorbing light. The leading thin film manufacturer uses a material that includes the rare element tellurium. Daniel Kammen, director of the Renewable and Appropriate Energy Laboratory at the University of California, Berkeley, says the presence of tellurium could limit the total electricity such cells could produce because of its rarity. While total worldwide electricity demand will likely reach dozens of terawatts (trillions of watts) in the coming decades, thin film solar cells will likely be limited to producing about 0.3 terawatts, according to a study he published last year. In contrast, the new cells from IBM could produce an order of magnitude more power.
The new cells could also have advantages compared to cells made of copper indium gallium and selenium (CIGS), which are just starting to come to market. That's because the indium and gallium in these cells is expensive, and while the selenium used in the IBM cell is rarer than indium or gallium, its cost is a tenth of either.
A new ink-based manufacturing process solves some of the key challenges to making efficient CZTS cells. A common approach to making any type of high-quality solar material is to dissolve a precursor substance in a solvent. This isn't possible with the CZTS cells because the zinc compounds required in the new cells aren't soluble. To get around this, the researchers used a combination of dissolved materials and suspended particles, creating a slurry-like ink that could then be spread over a surface that's been heat-treated to produce the final materials. The particles prevent the material from cracking and peeling as the solvent evaporates.
A semiconductor material could make solar cells three times as efficient.
Optical antennas could help solar cells produce more energy.
A new design can make more-efficient thin-film solar cells using existing manufacturing equipment.
Guest (AdrianMiller)
Thanks for the interest in this work! Although the authors haven't explicitly discussed how much their cells will cost, the materials are much more freely available than those used in current technologies, and the method for making them is also less difficult than that currently used.
If anyone is interested in knowing more about the science behind this article, we've set the original research paper free to access; you can find it here.
Adrian Miller
Advanced Materials
Mr. Miller,
Could you give me the detailed reference of the article in Advanced Materials, I woulb be interested to read it.
Many Thanks
Pierre Guichard
Guest (AdrianMiller)
The article is still currently in early view, and so doesn't yet have a full print reference. However, it is available online, either through the link I posted previously or through the DOI: 10.1002/adma.200904155. I hope that helps!
"the indium and gallium in these cells is expensive, and while the selenium used in the IBM cell is rarer than indium or gallium, its cost is a tenth of either"
Why is the selenium cheaper? And, if selenium is rare, how does the claim that the material is made from abundant elements make sense?
Thanks!
The cost has to do with both the abundance and demand. Right now there's a lot of demand for indium, for example, since it's used in displays.
I agree that the original sub-headline, with its emphasis on abundance is misleading. So we've changed it.
The new solar cell has a couple of main advantages. It does use more abundant materials than CdTe cells. It also avoids the use of some costly materials used in other types of solar cells. CIGS cells use selenium, but people are worried now about the indium, not the selenium, because of the big demand for indium from other tech. Right now, selenium supply isn't an issue. It would become an issue if the solar market increased by a few orders of magnitude.
But it would be really great is if they can get rid of that selenium entirely. The researchers have demonstrated that they can make thin films using the new process in which the selenium is completely replaced by sulfur. They haven't yet made solar cells with this film--that's in part because the research project is so new. They've only been working on it for 9 months. (That said, it could very well be that the selenium-free cells won't have very good efficiencies.)
Isn't selenium a significant component of agricultural runoff into California's Sacramento - San Joaquin River delta?
Replacing Selenium with Sulfur
Sulfur (S) and selenium (Se) are both similar to and different from each other. They are both Group (VI) elements and have similar chemical properties. Both are non-metals, but Se is closer to being a metal, meaning Se grips its outermost (valence) electrons less tightly than does S. These are the electrons that form chemical bonds with other elements and are also responsible for photoelectric properties. Se's bonds with metals (e.g., Zn) are more delocalized than the analogous bonds formed by S. So, for example, it’s not surprising to see that zinc sulfide (ZnS) has a wider band-gap (3.54 – 3.91eV, depending on the crystal structure) than ZnSe (2.82 eV). The extent of electron delocalization (and band-gap size) affects electrical, optical, and photovoltaic characteristics. The compounds mentioned in the article are more complicated than these simple sulfides, but I expect that substitution of Se with S has a significant effect on both the photovoltaic efficiency and light absorption characteristics. I doubt that a straightforward replacement of Se with S would yield a material with anywhere near equivalent properties. But they probably have strategies to compensate for this like varying the overall composition, crystal structure, and morphology.
It's good someone is taking seriously the opportunity to make excellent cells in a direct gap, inorganic semiconductor. That is the lesson of CIGS and CdTe, and worth others' time to act on.
Right now, even the best thin film, CdTe, is just barely ahead of Chinese-made multicrystalline silicon. That old dog has plenty of bite left in him, and will (along with CdTe) continue to make progress. $1/W Chinese c-Si at 15% is not far off, a daunting 'starting point' for any startup technology, as is $0.75/W 12% CdTe, which is also coming. People need to keep these near term, realistic goals in mind when they scale the wall of new technologies.
The tellurium issue is important, and studies are underway to help define it. The big levers are thinner layers (closer to what is needed to absorb the spectrum, not 3 microns) and higher efficiencies. The other big lever is more tellurium. BiTe mines with 17% tellurium ore exist, but the economical amounts are not well characterized; new byproduct Te will become available from PB, Au, Zn; and 9 million MT of the stuff exists in the well-known and well-studied undersea ferro-manganese ridges. Whether any can be extracted is another question.
This is a race to low cost PV, and the good news is, there is lots of reason to be optimistic that the race will be viciously competitive - with society being the beneficiary. Lowest cost installed systems are now less than solar thermal electric (about $3/W), and we will see $2/W within 5 years (probably), and down to the low $1 and change by 2030.
IBM is no frantic Silicon Valley start up needing to rationalize its existence with hype. Good luck to them! Their work is of value.
Manufacturing in the United States is in trouble. That's bad news not just for the country's economy but for the future of innovation.
Our list of the 50 most innovative companies, including the following:
On Twitter
Become a Fan on Facebook
Subscribe to the Feed
durs
44 Comments
Solar Cells from Cheaper Materials
So how much cheaper???
Cost is more important than efficiency.
Reply
gary7
59 Comments
Re: Solar Cells from Cheaper Materials
It's not cost VS efficiency. It's cost/watt of generating capacity. With that in mind, if you hold the production costs constant and increase the efficiency, the dollars/watts goes down or one could maintain the same efficiency and reduce manufacturing production costs and see the same improvement.
GAry 7
Reply