Novel type of molecules promises more efficient and cost-effective OLEDs displays

Printer-friendly versionSend by emailPDF version

Scientists from Harvard University have crafted more than 1,000 novel blue-light releasing molecules for organic light-emitting diodes, known as OLEDs that could drastically enhance displays for phones, tablets, televisions and more.


Screens of OLEDs utilize organic molecules that release light when an electrical current is applied. Unlike ubiquitous liquid crystal display that is LEDs, the OLED screens do not need a backlight, implying that the display can be as flexible and thin as a layer of plastic. Separate pixels can be switched off or turn on, drastically enhancing the colour contrast of the screen and consumption of energy. OLEDs are already substituting LCDs in high-end consumer gadgets, but a lack of efficient and stable blue substances has made them less competitive in big displays like televisions.

The interdisciplinary group of researchers from Harvard, in association with Samsung and MIT, introduced a big-scale, computer driven process of screening, known as the Molecular Space Shuttle that includes experimental and theoretical chemistry, chem-informatics and machine learning to rapidly discover novel OLED molecules that function as efficiently as industry standards.

The greatest challenge in crafting budgetary OLEDs is the release of the colour blue. Like LCDs, OLEDs rely on blue, green and red coloured subpixels to generate every type of colour on the screen. But it has been troubling to discover organic molecules that effectively release blue light. To enhance efficiency, OLED producers have generated organometallic molecules with costly transition metals such as iridium, to improve the molecule via phosphorescence. The solution is costly and yet still not deliver stable blue colour.

Aspuru-Guzik and his entire research team though to substitute organometallic systems with completely organic molecules. The team instigated building libraries of more than 1.5 million candidate molecules. After this, the range of molecules was narrowed down on the basis of predicting molecules that can deliver good outcomes. “It was a natural association between machine and chemistry learning,” says David Duvenaud, a co-author of the research paper and postdoctoral fellow in the Adams Lab.

According to Adams, “machine learning tools are truly advancement and starting to witness applications in a range of scientific domains. Such collaboration was an excellent opportunity to enhance the state of the art in computer science, while also introducing absolutely novel materials with practically feasible applications. It was extremely lucrative to witness such designs go from machine learning predictions to gadgets that you can take into your hand.”

“The overall success of such hard effort curtails from its multidisciplinary nature,” says Aspuru-Guzik. “Our associates at Samsung and MIT offered critical feedback about the needs for the molecular compositions.”

“The big throughput screening methodology pioneered by the team of Harvard researchers drastically diminished the requirement for synthesis, optimization, and experimental characterization,” says Marc Baldo, Lecturer of Electrical Engineering and Computer Science at the MIT. “It reveals to the industry the ways to advance OLED technology more efficiently and faster.” After such an enhanced design cycle, the group was left with numerous of molecules that function as well not much better than the state-of-the-art OLEDs free from metals. Applications of such sort of molecular screening also outspread far beyond OLEDs.