Reshaping the solar spectrum to turn light to electricity

July 28, 2015, University of California - Riverside
Photographs of upconversion in a cuvette containing cadmium selenide/rubrene mixture. The yellow spot is emission from the rubrene originating from (a) an unfocused continuous wave 800 nm laser with an intensity of 300 W/cm2. (b) a focused continuous wave 980 nm laser with an intensity of 2000 W/cm2. The photographs, taken with an iPhone 5, were not modified in any way. Credit: Zhiyuan Huang, UC Riverside.

When it comes to installing solar cells, labor cost and the cost of the land to house them constitute the bulk of the expense. The solar cells—made often of silicon or cadmium telluride—rarely cost more than 20 percent of the total cost. Solar energy could be made cheaper if less land had to be purchased to accommodate solar panels, best achieved if each solar cell could be coaxed to generate more power.

A huge gain in this direction has now been made by a team of chemists at the University of California, Riverside that has found an ingenious way to make conversion more efficient. The researchers report in Nano Letters that by combining inorganic semiconductor nanocrystals with organic molecules, they have succeeded in "upconverting" in the visible and near-infrared regions of the solar spectrum.

"The infrared region of the solar spectrum passes right through the photovoltaic materials that make up today's ," explained Christopher Bardeen, a professor of chemistry. The research was a collaborative effort between him and Ming Lee Tang, an assistant professor of chemistry. "This is lost, no matter how good your solar cell. The hybrid material we have come up with first captures two infrared photons that would normally pass right through a solar cell without being converted to electricity, then adds their energies together to make one higher energy photon. This upconverted photon is readily absorbed by photovoltaic cells, generating electricity from light that normally would be wasted."

Bardeen added that these materials are essentially "reshaping the solar spectrum" so that it better matches the photovoltaic materials used today in solar cells. The ability to utilize the infrared portion of the could boost solar photovoltaic efficiencies by 30 percent or more.

Photographs of upconversion in a cuvette containing (a) an optimized cadmium selenide /9-ACA/DPA and (b) a cadmium selenide /ODPA/DPA mixture. (9-ACA: 9-anthracenecarboxylic acid; ODPA: octadecylphosphonic acid; and DPA: 9,10-diphenylanthracene.) They were excited with a focused continuous wave 532-nm laser. The violet DPA output in (a) swamps the green beam that is clearly seen in (b), where no upconversion takes place. This indicates the enhancement of the upconverted fluorescence by the 9-ACA ligand. The photographs were taken with an iPhone 5 and were not modified in any way. Credit: Zhiyuan Huang, UC Riverside.

In their experiments, Bardeen and Tang worked with cadmium selenide and lead selenide semiconductor nanocrystals. The organic compounds they used to prepare the hybrids were diphenylanthracene and rubrene. The cadmium selenide nanocrystals could convert visible wavelengths to ultraviolet photons, while the lead selenide nanocrystals could convert near-infrared photons to visible photons.

In lab experiments, the researchers directed 980-nanometer infrared light at the hybrid material, which then generated upconverted orange/yellow fluorescent 550-nanometer light, almost doubling the energy of the incoming photons. The researchers were able to boost the upconversion process by up to three orders of magnitude by coating the cadmium selenide nanocrystals with organic ligands, providing a route to higher efficiencies.

"This 550—nanometer light can be absorbed by any solar cell material," Bardeen said. "The key to this research is the hybrid composite material—combining inorganic semiconductor nanoparticles with organic compounds. Organic compounds cannot absorb in the infrared but are good at combining two lower energy photons to a higher energy photon. By using a , the inorganic component absorbs two photons and passes their energy on to the organic component for combination. The then produce one high-energy photon. Put simply, the inorganics in the composite material take light in; the organics get light out."

Besides solar energy, the ability to upconvert two low energy photons into one high energy photon has potential applications in biological imaging, data storage and organic light-emitting diodes. Bardeen emphasized that the research could have wide-ranging implications.

"The ability to move light energy from one wavelength to another, more useful region, for example, from red to blue, can impact any technology that involves photons as inputs or outputs," he said.

Explore further: Towards more efficient solar cells

More information: Nano Letters,

Related Stories

Towards more efficient solar cells

August 13, 2014

A layer of silicon nanocrystals and erbium ions may help solar cells to extract more energy from the ultraviolet (UV, high-energy) part of the solar spectrum. Experimental physicists from the FOM Foundation, the STW Technology ...

Optical materials: Solar cells step up

November 20, 2013

The Sun is our most promising source of clean and renewable energy. The energy that reaches the Earth from the Sun in an hour is almost equivalent to that consumed by humans over a year. Solar cells can tap this massive source ...

Recommended for you


Adjust slider to filter visible comments by rank

Display comments: newest first

4 / 5 (4) Jul 28, 2015
Still doesn't make them work at night, so you have to factor in the huge cost of storage.
not rated yet Jul 28, 2015
I worked with frequency doubling crystals several years ago. IIRC they were extremely inefficient. We put them inside the laser cavity for just that reason. Has there been that much advancement recently?
3 / 5 (2) Jul 28, 2015
Why does everyone constantly bring up that solar doesn't work at night....

Please realize that current energy grids are pure demand only. There are no batteries. And thus we literally waste energy and dump it when we over produce which is all the time, because it is impossible to perfectly match supply with demand. So you over supply by just a little bit if your lucky every day. We should have created a battery system years ago. Super conducting batteries have a 95% efficiency and a 20% over head for cooling to storage ratio.

solar is still cheaper than dealing with the long term effects of coal , and natural gas. And if we invested in energy transmission from Europe, Africa, Australia and India a joint world project could end energy issues globally. Too bad no one wants to pay for that... and that would be expensive.

5 trillion dollars could solve world energy issues... right now
5 / 5 (1) Jul 28, 2015
We keep bringing it up because advocates of weather/sidereal-dependant renewables keep ignoring the cost of storage. At the moment it is the fossil fuel stations that duplicate capacity for when the renewables aren't available, which for solar is night and winter north of 50deg. The "long term effects of coal, and natural gas" are to provide plant food! The only sensible option for the medium-long term is nuclear. "5 trillion dollars" to transmit power around the planet with all the associated losses? Very wasteful. Spend it on nuclear where it's needed.
3 / 5 (3) Jul 31, 2015
Oh, GEEZ. Am I SERIOUSLY the only person at a site called "" that understands what the SOLAR CONSTANT is?

It's ONE kW per square METER folks.

With PERFECT solar cells and PERFECT weather, that is ALL you're going to EVER get from solar energy.

Now divide the generating capacity of the USA by 1kw and multiply by 1 sq.m. Now divide that number by a million and THAT -- a number in the THOUSANDS -- is the number of SQUARE KILOMETERS you're going to have to cover -- given PERFECT CELLS and PERFECT weather.

Now multiply that by FOUR, since that's the maximum number of hours -- SIX -- of SOLAR INSOLATION available anywhere in the USA. Now multiply AGAIN by about 2 to 3, that being the ACTUAL conversion rate you're going to get from REAL solar cells and not perfect ones -- 30 to 50%.

Now go tell Greenpeace or any nearby liberal how many THOUSANDS OF SQUARE KILOMETERS you're planning to cover with solar cells. Stand back, and watch their heads explode.
3 / 5 (2) Jul 31, 2015
Now go tell Greenpeace or any nearby liberal how many THOUSANDS OF SQUARE KILOMETERS you're planning to cover with solar cells. Stand back, and watch their heads explode.

SMART people will also note that I did NOT include any factors for storage and transmission losses. Not entirely trivial, **either**.
5 / 5 (1) Aug 19, 2015

Okay wise guy LETS DO THE MATHS
according to : http://www.eia.go..._02.html

2013 : power sold in US 3,725,101 MkWh
3,725,101,000,000,000 Wh - year = x
x / 365 = average per day = 10,205,756,164,384 Wh -- if we use storage = Y1
x / 365 / 24 = average per hour = 425,239,840,183 W -- if we use tradition energy source at night = Y2

Y1 * 1.3 for peak load variance = 13,267,483,013,699.2 = Z1
Y2 * 1.3 for peak load variance = 552,811,792,237.9 = Z2

I choose Sun Power X21-345 solar panels

I assume they use perfect conditions for testing; 345 W and a size of 1.6 m^2
Lets assume you only get 75% of this number and thus 259 W

5 / 5 (1) Aug 19, 2015
Z1 / 259 W = 51,225,803,142 panels ( 1 day of power ) = D1
Z2 / 259 = 2,134,408,465 panels ( 1 hr of power ) = H1

D1 * 1.6 m^2 = 81,961,285,027.2 m^2
H1 * 1.6 m^2 = 3,415,053,544 m^2

to get km^2 divide by 1 million

1 day of power needs 82,000 km^2 of land
1 hour of power needs 3,500 km^2 of land

Does not account for space between panels and access roads and building for storage and maintenance and all the other infrastructure -- so lets blow this up to 2x the number

160,000 km^2 for land if you choose to use batteries to store power for night usage
400 km ( 260 miles ) per side if square

7,000 km^2 of land if you switch to traditional sources of power for night time generation
83 km ( 52 miles ) per side if square

-- What to do about bad weather ?

Arizona is 300 miles x 400 miles so you can do say 4 H1 ( 4 hours of electricity ) setups across Arizona , New Mexico, and Cali

That is a TON of materials - but it's a doable infrastructure project
5 / 5 (1) Aug 19, 2015
so that 160,000 km^2 is if you are trying to generate all the energy you use in one day in just 1 hour and plan on storing it -- it is meant to be the over the top number.

Energy transmission is not addressed.

But 7000 km^2 is nothing... each state could do a 10 mile ^2 setup to reduce the over all load of transmission costs and share the power across the country. Except Alaska and Hawaii.

i almost forgot
HEY GREENPEACE -- WE NEED 14,000 - 28,000 km^2

you happy now?

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.