Portland, Ore. - Semiconducting polymers embedded with lead-sulphide nanocrystals could produce a light source for integrated photonic chips, according to recent work at the University of Toronto. The technique, producing infrared light at wavelengths used in communications systems, could be used to create photonic components orders of magnitude less expensive than current components, which can cost as much as $1,000.
Ted Sargent said his group in the university's Department of Electrical and Computer Engineering is also working on ways to build photonic crystal structures that could be deposited on silicon to create complete integrated photonic circuits. "We have found a way of making quantum dots of such a quality that they can produce light efficiently," said Sargent. He collaborated with professor Gregory Scholes' group in the Department of Chemistry. Scholes is excited about the potential efficiency of the quantum dots. Simulations by his group have shown that by using the third-order optical nonlinear response, light intensities could be 30 times that of gallium-arsenide devices.
"Our work plays a role in the conversion of energy from electrical to optical-we've shown a measurable efficiency at converting electrons into photons at wavelengths with which we can communicate," Lewis said. "This is where our work has relevance-at the interface between electronics and optics," he added.
Simpler, less expensive
Besides the nanocrystals' smaller size and, potentially, higher efficiency than today's electrical-to-optical converters, Sargent said his process for embedding them into a semiconducting polymer inherently simplifies and reduces the cost of optical chip manufacturing. Unlike traditional semiconductor fabrication techniques, the quantum dot and polymer fabrication techniques took place at normal pressures and relatively low temperatures.
By using simple thin-film techniques, Sargent sidesteps the exotic ovens and vacuum chambers needed to fabricate gallium-arsenide-style optics. As a result, his polymer/crystal nanocomposite produces light from electricity in a manner similar to an LED, but without the expense of traditional semiconductor fabrication.
Electrical measurements revealed an asymmetric, strongly superlinear current-voltage curve characteristic of a light-emitting diode, Sargent said, with internal efficiency of 1.2 percent.
"In a circuit sense, our quantum dots look a lot like forward-biased diodes. We apply a voltage bias, the current flows and the dots light up," he said. "Our objective here is efficiency, to have as much of that current as possible go into producing the light.
"We use nanocrystals of lead sulphide, a compound semiconductor unlike silicon, which is an elemental semiconductor," said Sargent. The team's cost-effective technique allowed it to manufacture the lead-sulphide nanocrystals at normal pressures and at the relatively low temperature of just under 150 degrees C. Today, more-expensive gallium-arsenide crystals convert electricity to light for fiber-optic communications and are fabricated in a vacuum at temperatures between 600 degrees and 800 degrees C.
After fabricating the lead-sulphide nanocrystals with wet-chemistry techniques, the group embedded the crystals into a semiconducting polymer, creating a polymer/crystal nanocomposite. By stabilizing the surfaces of the nanocrystals with a special layer of molecules (technically, a capping layer of oleate and octylamine ligands), the nanocomposite material becomes uniform enough to be spun onto a substrate as a thin film of hybrid polymer/crystal nanocomposite.
Unlike gallium arsenide-the compound semiconductor in today's optical chips-lead sulphide, Sargent said, does not need its formulas "tuned" (GaAs to InGaAs to InGaAsP) to produce the comm wavelengths. Instead, with nanoparticles, the material stays the same, the different wavelengths of light addressed by making different-sized dots (sizes ranged from just under 2 nanometers to just over 10).
"The great thing about dots, is that because they are so small, the energies that electrons are allowed to adopt inside them are determined by the diameter of the dot-so that a big dot produces long wavelengths of light, whereas short wavelengths are produced by smaller dots," he said.
Electron-hole pairs-excitons-combine in a quantum dot, releasing energy in the form of light. In a bulk semiconductor, excitons typically move freely, but when the exciton is trapped in a quantum dot with dimensions of the same order as the exciton, a confinement effect occurs, resulting in luminescence. By carefully picking the dimensions of the quantum dot, the frequency of the corresponding emitted light is precisely controlled.
"We are using electronic excitation to convert the energy we transferred to the dots into light, and the really important thing is that we've designed the dot so that it can address any wavelength between 1 and 2 micron," said Sargent. "1.3 to 1.6 micron is the band most often used by fiber-optic communications today, but the materials they use, such as gallium arsenide, have a much more restricted bandwidth, typically requiring different formulas for different wavelengths."
By running a current through the semiconducting polymer, the researchers were able to inject current through the polymer into the quantum dots, causing them to emit light. When the electrons encounter the semiconducting polymer, Sargent said, they must cross through the special coating to produce light. Consequently, Sargent tested many different coating methods before deciding on one that promotes a steady flow of electrons into the dots.
Despite making quantum dots produce light, Sargent can offer only educated guesses to explain the physical mechanism by which the quantum dots are excited. Electron and holes may be simultaneously tunneling to the dot, he suggests, or energy might be transferred without the physical movement of the electron-hole pair.
Regardless of what the physical mechanism turns out to be, Sargent's objective as an engineer is to make the energy conversion from electricity to light as efficient as possible. Consequently, he focuses his research on characterizing materials that result in the highest efficiency.
"We're not sure if it's tunneling or whether the electron and hole are just donating their energy, but we do know what materials to put around the dots to maximize the efficiency of the effect, and that's what really matters," said Sargent.
Sargent's ultimate goal is to create arrays of these devices operating in parallel so that macroscopic devices can be fabricated with them. For now, Sargent said, he will be satisfied with taking his proof of concept from here to a real silicon chip in "a small number of years."
"We have proven that we can put our material on a glass substrate, pass current through it and it emits light. Now what I want to do is show a chip itself-take an optical circuit that bridges the gap with electricity, and build it using our materials on a silicon substrate," Sargent said.
Sargent believes his lead-sulfide materials can be made to be compatible with traditional semiconductor substrates, for eventual integration with CMOS, but he cautions that it will take at least five years.