NREL Solar Cell Sets World Efficiency Record at 40.8%
13 August 2008
Scientists at the US Department of Energy’s National Renewable Energy Laboratory (NREL) have set a world record in solar cell efficiency with a photovoltaic device that converts 40.8% of the light that hits it into electricity. This is the highest confirmed efficiency of any photovoltaic device to date, according to NREL.
The inverted metamorphic triple-junction solar cell was designed, fabricated and independently measured at NREL. The 40.8% efficiency was measured under concentrated light of 326 suns. One sun is about the amount of light that typically hits Earth on a sunny day. The new cell is a candidate for the space satellite market and for terrestrial concentrated photovoltaic arrays, which use lenses or mirrors to focus sunlight onto the solar cells.
The new solar cell differs significantly from the previous record holder, which was also based on a NREL design. Instead of using a germanium wafer as the bottom junction of the device, the new design uses compositions of gallium indium phosphide and gallium indium arsenide to split the solar spectrum into three equal parts that are absorbed by each of the cell’s three junctions for higher potential efficiencies. This is accomplished by growing the solar cell on a gallium arsenide wafer, flipping it over, then removing the wafer. The resulting device is extremely thin and light and represents a new class of solar cells with advantages in performance, design, operation and cost.
NREL’s Mark Wanlass invented the original inverted cell, which recently won a R&D 100 award. His design was modified by a team led by John Geisz that further optimized the junction energies by making the middle junction metamorphic as well as the bottom junction. Metamorphic junctions are lattice mismatched—their atoms don’t line up. The material properties of the mismatched semiconductors allows for greater potential conversion of sunlight.
Hmm what the hell is "metamorphic cell"?
Posted by: yarross | 13 August 2008 at 03:35 PM
Quote " Metamorphic junctions are lattice mismatched—their atoms don’t line up. The material properties of the mismatched semiconductors allows for greater potential conversion of sunlight. "
From the above I'd guess that a metamorphic cell is one where the sunlight interacts with the solar cell to "enhance" the geometry and thus produce more power.
40.8 % is pretty sweet. I wish they'd make those in bulk shingles like the amorphous ones (ovonics).
Posted by: | 13 August 2008 at 05:40 PM
If sunlight is free why worry about efficiency. It is the cost of producing energy that is very important. This particular solar cell has a cost of several million dollars per kilowatt hour produced. It is very good to invent and build such things. They may save some on cost in orbit if the concentrator is not too heavy. I am not against expensive semiconductor art work that fossil fuels have made possible.
Parabolic mirrors that concentrate light upon stirling engine receivers may still be the most efficient way of producing electrical energy from sun light. ..HG..
Posted by: Henry Gibson | 13 August 2008 at 06:52 PM
40.8% is pretty nice, but since it requires 326 suns to achieve that efficiency, how efficient are the mirrors required to focus sunlight to those levels?
Can the entire system also be manufactured inexpensively and mass produced? If so, when do they expect to do so?
At 40% efficiency, if they can produce these at $3-5/watt that would be very good in today's market.
Posted by: Dave | 13 August 2008 at 06:56 PM
"If sunlight is free why worry about efficiency."
because land and space to build the panels is not free.
Posted by: philmcneal | 13 August 2008 at 07:29 PM
Love to see some good basic research because good things seem to come out of it one day. But I don't think this could have anything to do with real-world energy generation. Gallium Arsenide proved to be much faster and more efficient than silicon 25 years ago for making computer chips. Who cares....it's so expensive that even the military won't bother with it.
Oh, did I mention it's incredibly toxic? LOL
Posted by: DaveD | 13 August 2008 at 09:28 PM
I understand that, because these are so expensive per SQIN, they need to use concentrators. But as asked above, what is the cost per watt?
We can only assume that the cost per watt is still unreasonable.
We can also assume that the cooling of a cell with 326 suns (~3000 w/SQFT) is also daunting.
So, 40% is truly admirable and these guys deserve acclaim but, we just wonder, how close does it come to relating to the real world.
What is the cost per square foot, and the efficiency at one sun, we’re just curious.
Posted by: ToppaTom | 13 August 2008 at 10:04 PM
Regarding the system efficiencies: in such a concentrating system, mirrors could be over 90% efficient; yet other required components (such as homogenizer to prevent uneven illumination on the PV cells that destroys electric efficiency and harms cells) could reduce this to about 75% - depending on your specific design.
The cooling doesn't take much energy, and in addition, if you pay attention you can design it so that the removed heat is used as an energy product - for absorption chillers, water heating etc. Doing that can bring you to a total system efficiency in the 60-70% range.
I'm working in research of such systems at Tel Aviv University. It is very fast approaching a commercial product.
Posted by: Yosef | 13 August 2008 at 11:32 PM
Yosef: Sounds very interesting ... can you reveal more details? Do you have any numbers on the economic side of it?
Posted by: OldNeil | 14 August 2008 at 04:00 AM
@ Yosef
Please take a look at this link and tell us your opinion:
http://www.greencarcongress.com/2006/10/solar_systems_t.html
Posted by: Jorge | 14 August 2008 at 08:33 AM
If sunlight is free why worry about efficiency....
Parabolic mirrors that concentrate light upon stirling engine receivers may still be the most efficient way of producing electrical energy from sun light.
Sunlight is free, parabolic mirrors are not.
Let's say you need a $50k investment in mirrors, accompanying structures, tracking motors and so forth to concentrate 100 kW of raw solar energy. Consider cost per watt for:
1) 20% silicon wafers
($50,000 + $200) / 20 kW = $2.51/W
2) 30% Stirling engine
($50,000 + $10,000) / 30 kW = $2.00/W
3) 40% Mutltijunction PV
($50,000 + $20,000) / 40 kW = $1.75/W
"Why worry about efficiency?" Because efficiency drives concentrator price/kW.
By the way, I don't understand why 40.6% is a record. I thought Spectrolab did 41 or 42% a while back.
Posted by: doggydogworld | 14 August 2008 at 09:15 AM
One other cost is the tracking hardware and upkeep. Concentrators need to be pointed directly at the sun and that requires some expensive mechanical devices to do the job.
Posted by: sjc | 14 August 2008 at 11:36 AM
42.8 percent from July 23, 2008
http://www.udel.edu/PR/UDaily/2008/jul/solar072307.html
Posted by: abcd | 14 August 2008 at 12:07 PM
Looks like the scholars beat the feds.
Posted by: | 15 August 2008 at 11:14 PM
@OldNeil: There are numbers, but sadly I can't tell them. I can tell you that the exact benefit depends on how well you can match the heat consumption of a building (or storage) to what you can produce.
@Jorge: The interesting part there is the use of PV in a solar tower. The tower technology is probably already the most efficient for large scale (power plant) production, it's great if they manage to improve it even further.
Posted by: Yosef | 16 August 2008 at 11:24 PM
this is tremendous news. the more efficient it is, the cheaper it'll become and the more people will buy it. which in turn will make it cheaper still. a self perpetuating cycle
Posted by: Total Solar Energy | 19 September 2008 at 02:14 AM