Scientists make bendable, transparent LEDs—without organics

Organic LEDs, or OLEDs, promise to bring flexible, transparent displays to the market, but some researchers have found a way to get the same effect by printing microscopic inorganic LEDs onto plastic and glass.

Organic light emitting diodes, or OLEDs, promise to bring a great deal of flexibility to where we can put a display—literally. Because of their organic components, it should be possible to create flexible and transparent displays, opening up a large number of potential uses. But now, just as OLEDs may finally be ready for the consumer market, some engineers have figured out a way to get many of the same properties using inorganic LEDs (ILEDs), using a method that's so simple, even a biologist could understand it. It's a few years away—at least—from commercialization, but it's a significant advance.
The paper that describes the process will be published today in Science. The basic idea is that, since LEDs are so efficient at converting electrical charges to light, the human eye can detect the light of very small LEDs. As a result, it's possible to make a display out of a surface where only a small fraction is occupied by the actual LEDs, which can be small enough to be invisible to the naked eye. Under these conditions, the display will take on the properties of whatever material the LEDs are embedded in: bendable, transparent, etc.
Unfortunately, although we've gotten rather good at depositing the layered structure needed for making a normal ILED, the manufacturing processes we use don't scale down to the size of individual pixels in a typical display, which need to be on the order of 100µm or less. So the researchers came up with a simple solution: make a big one, and then chop it into little pieces.
The researchers started with a permanent substrate topped with aluminum arsenide, and layered on all the typical materials (gallium, indium, phosphorous, etc.) needed to create an LED that would glow red. When that was completed, they used a technique called plasma ion etching to cut a rectangular grid into the slab, leaving behind small squares, approximately 50µm across, held in place by the AlAs substrate. The squares were then anchored in place by a small bit of material in two corners, and the AlAs substrate was etched away with hydrofluoric acid. What was left was a grid of small LEDs held in place by two small posts that could be broken away easily.
This array, however, is packed so tightly that it would completely obscure any surface it was transferred to. So the authors crafted printing devices from a flexible material that only contains slots for a subset of the total LED square (say, every third one). The elastic material can pick up the LEDs, "print" them onto a separate surface, and then return to the original source and pick up the next set over. By adjusting how far apart the LED slots are—every second LED in the grid, or every fifth—it's possible to print out devices with different spacing.
The authors prepared a flexible plastic surface by laying down a grid of wiring, printed the LEDs on it, and then locked them in place with epoxy; a second mesh of wiring completed the circuit. Given an adhesive, the plastic could be applied to just about anything. For demonstration purposes, the authors stuck it to a glass cylinder. They also created a wiring grid that acted as a passive matrix, allowing them to activate individual LEDs in the grid. For these applications, the LEDs only took up about one percent of the total surface, enough to leave it transparent (provided they weren't lit, obviously).
The authors also demonstrated how to create a bendable display. By pre-stretching the flexible plastic substrate before laying down the wiring, the wiring would buckle within the material when tension was relaxed. This provided enough slack to accommodate a fair degree of flexibility in the final material.
If, at this point, you think you're missing something, you're not—it really is that simple.
That said, it's still a long way from being ready for the market. The authors say that none of the LEDs failed during testing, but some of the wiring leading into the device wasn't up to the strains of their test procedures. As a result, most of the individual devices they made had rows of dead pixels. All the devices only worked in red, as well. Still, the process uses well-understood materials and techniques, so there's no reason it can't be rapidly improved on, and the transition to a production environment doesn't seem to face any major stumbling blocks.

Original post is here.