Researchers at ETH Zurich have made light-emitting polymers that survive being processed like a photoresist, and used them to pattern an OLED test image at roughly a micrometre per pixel. The work was published in Nature on 16 September and announced by the university on 23 September.

Why lithography and OLEDs do not normally mix

Photolithography is the step the chip industry has spent five decades learning to do precisely, but it is hostile to organic emitters. Patterning a film that way means exposing it to ultraviolet light and then to developers and solvents, and the conjugated molecules that make an OLED emit tend not to come through that chemistry intact. Production lines avoid the problem instead of solving it: red, green and blue emitters are evaporated through fine metal masks, or deposited from solution by inkjet, and both routes put a floor under how small a subpixel can be.

The ETH group's answer is to separate the two jobs the material has to do. Their emitters use a core-shell architecture: the light-emitting unit sits at the centre, wrapped in a star of arms whose outer tips carry the groups that react to UV. Exposure cross-links those tips to their neighbours, making the illuminated areas insoluble, while the inner part of the shell keeps the developer away from the emitter itself. What is left behind after development is a patterned, electroluminescent film.

“We separate the two functions spatially,” said Chih-Jen Shih, a professor at ETH Zurich and one of the paper's senior authors. “The light-emitting molecule is protected inside, whilst the reactive cross-linking groups are on the outside.”

What they actually built

Two demonstrations are reported. The first is a multicolour fluorescent image of a parrot measuring 300 by 430 micrometres and made up of 250 by 350 pixels — which works out at about 1.2 micrometres per pixel, and which ETH describes as the highest-resolution multicolour fluorescent image produced by photolithography so far. The second is an electrically driven device: a glowing ETH logo, 1 by 2.4 millimetres, showing that the patterned polymer still works as an OLED once contacted.

It is worth being clear about what is not in the announcement. Neither ETH nor the paper's summary reports external quantum efficiency, luminance or operating lifetime for the lithographically defined devices, and those are the numbers that decide whether an emitter is manufacturable rather than merely patternable. The parrot is a fluorescent image, not a working display; the logo is a working device, but a single large one rather than an addressable array.

Where it could go

The team's stated next steps are to shrink the emitting pixels further and to build the per-pixel drive electronics that a display would need — the harder half of the problem, and one that pushes the work towards silicon backplanes rather than the plastic substrates this publication usually covers. ETH also points at medical imaging, neuroscience, microscopy and sensing, all applications where a patterned light source on a chip is useful and where display-grade lifetime matters less.

For the printed-electronics industry the interest is less in microdisplays than in the principle. A patterning chemistry that leaves an organic semiconductor functional widens what can be done with organic materials generally, and it arrives while the display supply chain is still working through the cost of the incumbent methods — the materials bills that keep rising and the flat shipment forecasts that make new capital equipment a hard sell. It also sits alongside other recent work putting optoelectronic function onto wafers that were not supposed to carry it.

The paper is credited to seventeen authors across ETH Zurich, RMIT University in Melbourne and other institutions, with Yinyin Bao — formerly at ETH Zurich and now a professor at the University of Helsinki — among the senior authors alongside Shih and ETH professor of electronics Hua Wang.