By Saren Johnston
Simple polymers are long chains of molecules, generally carbon with hydrogen on the sides, all attached to one another by single bonds. "That's the backbone," says Joe Shinar. "As you go down a chain, regardless of what is on either side, the bonds are all single bonds. Simple polymers are all very insulating, and the most common one is polyethylene."
But common isn't where the action's at for Shinar, an Ames Lab senior physicist and professor of physics at Iowa State University (ISU). He's going after a more complex "beast," his name for a group of materials called pi-conjugated polymers. He and members of his research group are investigating the physics of these materials in an effort to determine their potential for improving the performance and efficiency and reducing the cost of light-emitting diodes (LEDs) -- those wonderful semiconducting devices that when voltage is applied luminesce and make possible the gleaming displays of alphanumeric symbols seen on such familiar items as clocks, VCRs, microwave ovens and automobile dashboard panels, and that may one day replace the liquid crystals in flat panel displays.
Although pi-conjugated polymer sounds like a terribly technical term, the "beasties" are not as difficult to understand as their awesome appellation might lead us to believe. "Unlike a simple polymer, what you have in a pi-conjugated polymer as you go down the chain is alternating single and double or single and triple bonds," Shinar explains. "The energy of the electrons in both double and triple bonds is much higher than the energy of the electrons in single bonds. The energy gap, which is very large for an insulator, is much smaller for a material that has single-double or single-triple bonds, in other words, a pi-conjugated polymer. Virtually all pi-conjugated polymers are semiconductors."
Shinar continues, "Once you realize pi-conjugated polymers are semiconductors, the obvious question is, 'Hey, can we make this whole universe of devices, which we normally do with silicon, gallium arsenide, or other inorganic semiconductors, with pi-conjugated polymers?'"
The answer is "yes." Using pi-conjugated polymers for the emissive layer in LEDs, researchers can produce efficient devices with the full rainbow of visible colors. The emissive polymer layer is sandwiched between two dissimilar metals, serving as cathode and anode, on a glass substrate. The current injected from the cathode into the polymer takes the form of negatively charged particles, while the current injected from the anode consists of positively charged particles, which travel through the polymer until they encounter the negative particles. When the two meet, the recombination creates excited states called singlet or triplet excitrons. Some of the singlet excitrons decay to the ground state by nonradiative processes. But some decay radiatively, emitting the coveted light. By studying the decay mechanisms, Shinar and his co-workers hope to discover how to make pi-conjugated LEDs live long enough to be useful for commercial applications.
Contributing to that effort are Shinar's collaborators on the LED research. A group led by Tom Barton, Ames Lab director and professor of chemistry at ISU, has synthesized many of the polymers Shinar has investigated. And work has also been done in cooperation with other researchers around the world, most notably Z.V. Vardany and his group at the University of Utah.
But the overriding issue remains the same. "Stability is the critical factor," says Shinar, "whether it's light-emitting diodes, field-effect transistors, photovoltaic cells or even lasers. All of these devices have been made with pi-conjugated materials, but none of them live long enough."
There are some cases where Shinar says improvements have brought these pi-conjugated materials-based LEDs "tantalizingly close to the commercial requirements for stability. The most stable of the small pi-conjugated molecular devices fabricated to date are green LEDs based on materials developed at Kodak and now licensed to Pioneer. Pioneer has managed to tweak their devices to the point where they can run them for more than ten thousand hours," he says.
"Once you achieve the commercial stability requirements and start thinking about potential applications for these materials, then the advantages of pi-conjugated polymer LEDs become immediately obvious," Shinar continues. "Generally, you could make gigantic devices. If you wish, you could make a display where each segment is one meter square because fabrication of the device per unit area would be very cheap. Upscaling would be relatively much less of a problem than in any inorganic material, and because polymers are inherently flexible, they can be used to manufacture flexible LEDs. The applications are there; you just need to find out why the LEDs made with pi-conjugated materials die and what you can do to get over that big hurdle."
Even if scientists overcome the
stability problem, they must still solve the problem of how to
increase luminescence in pi-conjugated polymers. So, how do you
improve efficiency? Shinar says you start by asking, "What
is it that competes with the luminescence coming from these
materials? You're generating these excited states that luminesce,
so why don't they luminesce with a yield of 50 percent? Why do
they luminesce with a yield of less than 5 percent?"
To help answer these questions, Shinar and members of his research group make novel use of optically detected magnetic resonance (ODMR), a tool that is more typically applied to biological systems than to the physical systems associated with condensed matter physics. "We were quite simply the pioneers in applying ODMR to pi-conjugated materials," he says. "Not only were we the first, for about five years or so we were the only ones."
Magnetic resonance is a phenomenon exhibited by the nuclear or electronic magnetic moments. In a direct current (dc) magnetic field, these moments, or spins, absorb energy when they are subjected to magnetic fields alternating at the resonate frequencies that correspond to the difference between their energy levels. Because many of the nonradiative decay processes that either contribute to or compete with luminescence are spin-dependent, ODMR provides a magnetic resonance spectrum by monitoring the change in the optical quantity (e.g., luminescence) as a function of the dc magnetic field.
Within the last few years, interest in ODMR for physical systems has caught on among other research groups. "They decided, 'Hey, this is a very fascinating tool; we want to play with it ourselves'," says Shinar. "So now there are about three or four groups in the world doing ODMR on polymers and organics."
Proponents of ODMR include Richard Friend, Cavendish Professor at Cambridge University in England. "Richard Friend has made, at this point, the most important contribution in this area because his group fabricated the first polymer LEDs back in 1990," says Shinar. "I was chatting with him at the 1996 fall meeting of the Materials Research Society, and he said, 'You know, I think that ODMR is possibly the most stunning piece of physics done on polymer LEDs.'
"ODMR can really help in probing the microscopic processes that either help the luminescence or hurt it," says Shinar. "The information it provides will enable us to synthesize new materials that will weaken the nonradiative processes and, consequently, enhance the radiative decay needed to produce light."
Running ODMR experiments on different pi-conjugated polymers is the job of Ames Lab postdoctoral fellow Jonathan Partee and graduate student Brian Uhlhorn. Working in a darkened lab, illuminated by only a computer terminal and a laser beam, the two are flipping the spins of electrons and watching for excitrons that luminesce as they decay. In a typical ODMR experiment, Partee and Uhlhorn first place the sample polymer in a microwave cavity in a magnetic field. At the same time that microwaves are traveling to the sample, the two researchers direct an argon laser beam into the microwave cavity to photoexcite the sample, generating luminescing excitrons. The magnetic field is then slowly scanned over a range of values while the luminescence is monitored for any changes induced by the the microwaves, which supply the alternating magnetic field. Plotting the change in the luminescence versus the magnetic field gives an ODMR spectrum, which helps Shinar, Partee and Uhlhorn better understand the nonradiative decay processes and how they affect luminescence.
"Of course, there are other ways to assess nonradiative decay," says Shinar. "You can do something called photo-induced absorption spectroscopy in which you compare the absorption property of the material when it absorbs light from the ground state versus when it absorbs light from an excited state. This measurement, in combination with others, provides information on the excited states, such as their energy, whether they are charged or not, whether they have spin or not, etc."
Regular luminescence spectroscopy is another method of determining nonradiative decay. In this technique, researchers monitor the luminescence of the material in solid versus solution either as a function of concentration in solution, as a function of various side groups that can be attached to the main backbone of a polymer or as a function of changes that are made to the backbone.
"The most powerful use of any of these assessment techniques is in combination with one another," says Shinar. "ODMR provides beautiful complementary information, especially when compared to the photo-induced absorption, because most of the more revealing photo-induced absorption studies are picosecond, ultrafast. ODMR probes the excited states that are much more longer lived -- from say the one microsecond up to the 50 millisecond range."
Constructed like a simple sandwich, the pi-conjugated polymer LED consists of a thin, transparent conducting anode layer of indium tin oxide (ITO) deposited on glass, the organic polymer layer and an aluminum cathode. Shinar and his co-workers developed etching procedures for improving the interface regions between the ITO anode and the organic layer and between the organic layer and the aluminum cathode while working on a project with Dow Chemical.
"Initially, we thought what we were doing with the etching was turning a smooth ITO layer into a corrugated, hills and valleys and canyons kind of thing, which increases the contact area so you'll have the easiest injection of the greatest number of charge carriers from the ITO to the organic layer," he explains.
But Shinar and his co-workers soon discovered that the etching procedure was not as simple as it seemed. "It turned out that optimal etching is not the same for all organics," he says. "So there's more involved in that process, but what we did show is that you need to determine the optimal etching for any organic LED that you make.
"We managed to enhance the performance of our devices with these etching techniques by a hundred fold," says Shinar. "We fabricated efficient, bright blue LEDs that lived for about 50 hours. While the efficiency was high, the lifetime was insufficient," he admits. "But the developments we made in improving the interfaces between the electrodes on the LEDs should help any device, including those that last for thousands of hours."
To enhance the outlook for pi-conjugated molecular and polymer LEDs, Shinar and his colleagues know they must continue to revisit the old problems of efficiency and stability to increase luminescence and make the devices last longer. That means ODMR will continue to play a key role in their research.
"ODMR has already revealed much of the nature of the excited states of these materials, and some of them are actually much worse than we previously thought," says Shinar. "They live from 10 microseconds to 10 milliseconds, or about 10 thousand to 10 million times longer than the luminescent states. At the same time, the short-lived luminescent states are highly diffusive, and when they approach one of these long-lived 'bad states,' they are nonradiatively quenched back to the ground state. Hence, during the long life of the bad states, they 'kill' many of the luminescent states. Therefore, we must minimize their population," he explains.
By using ODMR capabilities, Shinar has also found indications of certain sites within pi-conjugated polymer materials that stabilize the long-lived quenching states, prolonging their lives. "We don't exactly know the nature of these sites, but our ODMR is indicating that they are ones where you have interchain coupling, neighboring chains in a certain configuration relative to one another," he says. "If we were able to identify these sites much more specifically, we could use that knowledge to engineer the films and LEDs in a way that would minimize these types of sites and hopefully give us much more efficient and perhaps longer-lived LEDs."
For more information:
Joe Shinar, 515-294-8706, shinar@ameslab.gov
Current research funded by:
DOE Basic Energy Sciences Office
National Science Foundation
Return to Inquiry table of contents
Last revision: 4/17/98 sd
Home | Comments | Search | Disclaimer