Study Concludes “Peak Coal” Will Occur Close to 2011
New Model Determines Fossil CO2 Emissions Need To Be Reduced by 56% By 2050 and Approach Zero by 2100 to Keep Warming No More Than 2 Degrees

New Solar Energy Conversion Process Could Boost Efficiency

Stanford engineers have developed a process—photon-enhanced thermionic emission, PETE—that simultaneously uses the light and heat of the sun to generate electricity in a way that could make solar power production more than twice as efficient as existing methods.

(a) Energy diagram of the PETE process. Photoexcitation increases the conduction-band population, leading to larger thermionic currents and enabling the device to harvest both photon and heat energy. (b) One possible implementation. Schwede et al. Click to enlarge.

Unlike photovoltaic technology currently used in solar panels—which becomes less efficient as the temperature rises—he new process excels at higher temperatures.

This is really a conceptual breakthrough, a new energy conversion process, not just a new material or a slightly different tweak. It is actually something fundamentally different about how you can harvest energy.

—Dr. Nick Melosh

Melosh is an assistant professor of materials science and engineering, and is senior author of a paper describing the tests the researchers conducted, published online 1 August in Nature Materials.

Most photovoltaic cells use the semiconducting material silicon to convert the energy from photons of light to electricity. But the cells can only use a portion of the light spectrum, with the rest just generating heat. This heat from unused sunlight and inefficiencies in the cells themselves account for a loss of more than 50% of the initial solar energy reaching the cell.

If this wasted heat energy could somehow be harvested, solar cells could be much more efficient. The problem has been that high temperatures are necessary to power heat-based conversion systems, yet solar cell efficiency rapidly decreases at higher temperatures.

Melosh’s group figured out that by coating a piece of semiconducting material with a thin layer of the metal cesium, it made the material able to use both light and heat to generate electricity. While most silicon solar cells have been rendered inert by the time the temperature reaches 100 °C, the PETE device doesn’t hit peak efficiency until it is well over 200 °C.

Because PETE performs best at temperatures well in excess of what a rooftop solar panel would reach, the devices will work best in solar concentrators such as parabolic dishes, which can get as hot as 800 °C. Dishes are used in large solar farms similar to those proposed for the Mojave Desert in southern California and usually include a thermal conversion mechanism as part of their design, which offers another opportunity for PETE to help generate electricity, as well as minimizing costs by meshing with existing technology.

Melosh calculates the PETE process can get to 50% efficiency or more under solar concentration, but if combined with a thermal conversion cycle, could theoretically reach 55 or even 60%—almost triple the efficiency of existing systems.

The researchers used a gallium nitride semiconductor in the proof of concept tests. The efficiency they achieved in their testing was well below what they have calculated PETE’s potential efficiency to be, which they had anticipated. But they used gallium nitride because it was the only material that had shown indications of being able to withstand the high temperature range they were interested in and still have the PETE process occur.

With the right material—most likely a semiconductor such as gallium arsenide—the actual efficiency of the process could reach up to the 50 or 60 percent the researchers have calculated. They are already exploring other materials that might work.

Another advantage of the PETE system is that by using it in solar concentrators, the amount of semiconductor material needed for a device is quite small, thereby helping to keep costs down.

The PETE process could really give the feasibility of solar power a big boost. Even if we don’t achieve perfect efficiency, let’s say we give a 10 percent boost to the efficiency of solar conversion, going from 20 percent efficiency to 30 percent, that is still a 50 percent increase overall.

—Nick Melosh


  • Jared W. Schwede, Igor Bargatin, Daniel C. Riley, Brian E. Hardin, Samuel J. Rosenthal, Yun Sun, Felix Schmitt, Piero Pianetta, Roger T. Howe, Zhi-Xun Shen & Nicholas A. Melosh (2010) Photon-enhanced thermionic emission for solar concentrator systems. Nature Materials doi: 10.1038/nmat2814



Wow...very promising research!




"Melosh calculates the PETE process can get to 50% efficiency or more under solar concentration, but if combined with a thermal conversion cycle, could theoretically reach 55 or even 60%—almost triple the efficiency of existing systems."

Very interesting. And to think they're using gallium arsenide - a building block of the very first transistors. Solar such as this will be a way to mitigate the impact of other, more disruptive technologies that require a completely new understanding of physics.

A test project set up in the desert would be the immediate goal for this.


They used GaN, not GaAs, but recommended that GaAs would be better.
"The researchers used a gallium nitride semiconductor in the proof of concept tests."
I'll have to read the paper.


This is good stuff. With present concentrators, they could convert heat and light, all you need is a cold mirror. This would allow them to do this at a lower cost, which is good.


Lower cost with higher efficiency is an interesting combination. @30+% efficiency, roof mounted collectors could supply most of th e-energy required for a residence + one EV. E-storage may still be a problem to solve but it will come.


Oh please! Harvey, your average single family suburban sprawl house has more than enough roof area to do that with just 10% efficient PV. You don't necessarily want the most powerful panels on your roof because the panels not only protect the roof but also lower the ambient temperature as they convert sunlight into electric so it's often better to use a greater number of low power panels to cover and insulate more of the roof than fewer of the higher power panels.


This is more for the solar field of a million watts per acre. I could use 4 kW for my house and 4 kW for my car, so I might need the most I could get from the south facing roof space that I have, not to mention wanting lower cost per watt.


And this process gains thermal efficiency at +200°C. Roof panels without concentrators lose the efficiency increase.


Why would you need 4kw for your car?
At 15% capacity that gives you around 5,500 Watt/hours year, enough for 2 Leaf cars at 12,000 miles/year.


At 5 hours per day generating 20 kWh I could go 80 miles in my car, a 40 mile commute each way.


Cheapness is what really matters when it comes to solar.

Whoever makes the lowest dollars-per-watt panel will win all the big prizes. Firstsolar and Nanosolar are well aware of this.


Yes, but like Unisolar, they are more likely to sell at market prices and use the rest for expansion and dividends.


The disadvantage is that high concentrations require clear skies. This will be suitable mostly in arid zones, not where clouds or haze are frequent.


The nice part is 100C can run absorption cooling. So with these on the roof, you could heat water, heat your home, cool your home, charge your car and provide electricity on the grid.

Henry Gibson

only the very rich people of the world have a large roof area. ..HG..

If you have one you are very rich. The poor people are sleeping now instead of writing comments.


The per-unit roof area of a 3-story apartment building is sufficient to supply the needs of the people who live there. Given the efficiency of the photoelectro-voltaic scheme here, it would be even easier.

The comments to this entry are closed.