Associate Professor Silvija Gradečak is developing nanowires for solar cells and light emitting diodes.
The next generation of solar cells may be flexible, transparent and more energy efficient, says Silvija Gradečak, the Thomas Lord Associate Professor in Materials Science and Engineering, whose Laboratory for Nanophotonics and Electronics at MIT is working to develop semiconducting nanowires for solar cells, as well as for light emitting diodes (LEDs) that can replace inefficient light bulbs.
Associate Professor Silvija Gradečak explains cathodoluminesence apparatus attached to a transmission electron microscope, which allows researchers to characterize the optical properties of a material with nano scale resolution by shining an electron beam onto the sample. Photo: Denis Paiste, Materials Processing Center
“Nanostructured materials would enable development of solar cells that are flexible, can be produced in large scale using roll to roll processing, and are potentially transparent, meaning that we could use them on surfaces like windows, cars, etc. The new class of nanostructured solar cells is not necessarily competing with the existing silicon technology, but it would enable development of devices that do not exist on the market right now. The goal is to develop solar cells that absorb as much solar light as possible and at the same time, we are developing LEDs that are producing light in as efficient a manner as possible,” Gradečak said.
Gradečak’s lab focuses on three pillars: growth, characterization and devices. “Nanostructured materials are unique because of their dimensions, and because of that nano dimensionality, they have unique properties that are very distinct from their bulk counterparts. The main challenge remains how to modify their structure, the chemistry, the doping on the atomic scale for a desired application. The main research goals in my group are to develop the growth techniques and methods that will allow us to modify the structure and the chemistry of nanomaterials on the atomic scale. To accomplish that goal, we combine different synthesis methods, characterization and device fabrication,” she said.
“Our goal is to understand the complexity of the nanowire synthesis that will enable us to modify the materials in a predefined way,” Gradečak said. “Nanowires themselves can act both as the absorption medium, meaning that they can absorb the light, and at the same time, they can transport the charges to the corresponding electrodes because of their one dimensionality.”
Graduate student Samuel C. Crawford works in Gradečak’s lab on nanowire growth with the goal to control their composition and morphology, or geometric shape. Crawford’s research has demonstrated that nanowire morphology can be controlled in situ, during the growth. (See related article and video.) This is particularly important and challenging because nanowires can be on the size scale of proteins, ranging from 10s to 100s of nanometers in diameter with a typical length range from 10s to 100s of micrometers.
Nanowire solar cells
In a Feb. 9, 2011, paper in Nano Letters, Gradečak and fellow MIT faculty member Vladimir Bulovic demonstrated 2.36 percent power conversion efficiency in a solar cell based on a blend of poly(3-hexylthiophene) (P3HT) and narrow bandgap gallium arsenide (GaAs) nanowires. The work was supported by Eni SpA under the Eni-MIT Alliance Solar Frontiers Program.
ZnO nanowire/PbS QD hybrid solar cell (left) and ZnO nanowire/P3HT hybrid solar cell (right). Images: Sehoon Chang
Several subsequent papers, including one with Bulovic, Moungi Bawendi, Mildred S. Dresselhaus and Jing Kong, published Dec. 3, 2012, demonstrated power conversion efficiencies of 4.2 percent in solar cells with zinc oxide (ZnO) nanowires grown on graphene paired with lead selenide quantum dots, while also showing 5.1 percent efficiency in solar cells on indium tin oxide (ITO) electrodes. Results were obtained under a standard lighting test, AM 1.5G, which simulates averaged terrestrial sunlight conditions. The results constituted the first demonstration of graphene-based hybrid photovoltaic devices using ZnO nanowires.
Graphene electrodes have the potential to replace ITO. “The indium used in ITO is expensive, while graphene is made from ubiquitous carbon, which is cheaper. It also provides other advantages, including flexibility, low weight, mechanical strength, and chemical robustness,” Gradečak wrote in an article for the Society of Photo-Optical Instrumentation Engineers (SPIE).
Gradečak and collaborators have two patents on nanowire synthesis using gold seed particles and electron beam lithography as well as two additional patent applications in process. “The patented inventions that have emerged from this work have been implemented, and we do have working devices,” she said. “I believe that these type of devices will be explored for real world applications.”
Gradečak’s group is currently also in the process of making LEDs that are based on their nanowires. “The global aim is to replace less efficient sources of artificial lighting, including incandescent light bulbs and fluorescent lamps, with so-called sold-state lighting devices that would be more efficient. This is because most of the electrical energy that is consumed by the device would emit light rather than produce heat as it is the case in conventional light sources,” Gradečak said.
Commercially available LEDs are relatively expensive. A 60-watt incandescent light bulb can be bought for about 62 cents, but its LED equivalent costs about $12, making it nearly 20 times more expensive. However, the LED bulb has a life expectancy of 22 years and uses 84 percent less energy than the incandescent bulb, which will last for only about 2,000 hours.
MIT News Office
Current LEDs based on materials such as gallium nitride are expensive because of the process used to create the gallium nitride, which is typically grown on a substrate of industrial grade sapphire or silicon carbide. “The reason why gallium nitride film has to be grown on these type of substrates is because of lattice matching...and these substrates are relatively expensive. Consequently, light emitting diodes that are based on gallium nitride, which is the material of choice for this application, are currently expensive. By using nanowires, we can overcome some of these issues,” Gradečak said.
“For nanowires, the lattice matching, or finding a substrate with similar crystallographic structure and lattice parameters, is not critical. For example, we routinely grow nanowires on lattice-mismatched substrates like silicon. We have demonstrated that we can grow nanowires on flexible substrates, on polymers, and on a variety of other substrates,” she said.
Besides its high cost, sapphire also resists transmitting heat. “It doesn't conduct heat very well, so the heat that is produced by LEDs has to be somehow dissipated, and sapphire doesn’t do a good job in that regard,” Gradečak said. Typically, the sapphire substrate will remain part of the LED, adding to its cost. Other researchers are investigating efficient ways of removing the gallium nitride-based LEDs from the sapphire substrate.
|Associate Professor Silvija Gradečak with a metal organic chemical vapor deposition system used to create nanowires from metal seed particles, such as gold, and gases flowing in a quartz chamber at high temperatures between 500 and 1,000 degrees Celsius. Elements, such as gallium, are delivered by gaseous precursors and alloy with the gold seed particle and then react to form one-dimensional pillars, such as gallium nitride. The composition and diameter can be changed along its length to alter its material properties. Photo: Denis Paiste, Materials Processing Center|
“In general, nanowires can be grown on a variety of substrates, so we can address the cost; at the same time, we can address the question of defects in gallium nitride material,” Gradečak said.
The Nanowire Advantage
In their bulk form, gallium nitride thin films, even when grown on substrates that have relatively similar lattice parameters, have relatively high densities of threading dislocations and other defects. “These type of defects quench the light – or the technical term is they act as nonradiative recombination centers, meaning that the electrons and holes recombine without producing light at these defects, so the efficiency of the light emission is reduced.
As it turns out, nanowires, in particular gallium nitride (GaN) nanowires, are dislocation free, so these are single crystalline materials that do not contain dislocations, and from the quality of the materials perspective, there are many advantages of using gallium nitride nanowires for these type of applications,” Gradečak said.
To advance fundamental understanding of the nanowire properties, which are grown by chemical vapor deposition at temperatures of 500 to 1,000 degrees Celsius, Gradečak’s group uses advanced electron microscopy techniques, including cathodoluminesence.
Cathodoluminescence allows researchers to study optical properties of materials with nanoscale resolution by shining an electron beam onto the sample. Coupled with conventional transmission electron microscopy techniques, cathodoluminescence lets researchers simultaneously observe the structure and the chemistry of the material as well as its optical properties. “In that way, we have the direct correlation between the structural and the optical properties of the material with unprecedented spatial resolution,” Gradečak said.