The new theory emerged from a joint project between scientists at the University of Mons-Hainaut in Belgium, the Center for Molecular Science at the Chinese Academy of Sciences (Beijing) and the School of Chemistry and Biochemistry at the Georgia Institute of Technology (Atlanta). The researchers worked closely with experimentalists hoping to identify new polymer species that would be competitive with other light-emitting materials. "The current work with polymer LEDs has been with the monochrome displays, but there has been much work on developing full-color displays, and there are prototypes that I have been told are really beautiful," said Jean-Luc Bredas who works at Georgia Tech. "With a lot of companies working in the area, plastic electronics is really coming."

Polymer light-emitting devices came out of initial research into conducting polymers starting with the work of Alan Heeger at the University of California at Santa Barbara for which he eventually received a Nobel Prize. Although compound semiconductors and organic crystalline semiconductors perform much better, it is polymers that can be inserted into low-cost, high-throughput manufacturing processes. But they are at the low end of the scale in terms of performance. Inorganic and organic semiconductors have a theoretical limit of 100 percent efficiency. In practice, such limits are never reached, but the figure at least holds out the promise that these compounds can be indefinitely improved. By contrast, polymers have a theoretical limit of only 25 percent, a disappointing prognosis for display engineering. Bredas and his colleagues began looking at how charge carriers — holes and electrons injected into the polymer at electrodes — recombine to produce light. Polymer chains form highways that conduct both types of charge to common meeting places where they organize themselves into short-lived pairs called excitons. Depending on their spin orientation, the carriers can form two different types of exciton, only one of which produces a photon when it decays. The other type, called a triplet exciton, dissipates its energy in the form atomic vibrations when it decays. The spin statistics are stacked against photon production since the light producing singlet excitons have a probability of formation of only1/4. That leads to the 25 percent efficiency limit. "These compounds consist only of light elements, carbon nitrogen hydrogen and oxygen, but what is needed for efficient conversion is heavy atoms," he explained. Heavier atoms can influence the proportion of singlet excitons produced, boosting photon production. In general, the heavier the atoms, the more difficult it becomes to process the materials, so the best performers are the least cooperative at the manufacturing stage. Crystalline organic semiconductors can be doped with heavier atoms while their light-weight organic nature makes them easier to process, although they still require vacuum deposition. In the new look at polymers, Bredas performed calculations on long-chain polymers to look at the dynamics of exciton formation.

"Triplet excitons take longer to reach their ground state than singletons, which gives them extra time to transform into photon-producing excitions," he explained. Triplets can dissociate and reform as singletons, and they can also convert via a process known as intersystem crossing.

Calculations show that the longer-chain systems favor the formation of singletons, indicating that there might be polymers that produce more of them than would be predicted by statistics alone. The theoretical studies are now being used in conjunction with experimental work to identify possible candidates with higher conversion efficiencies.