QnAs with Stephen R. Forrest

The smartphone revolution capitalized on generations of advancements in inorganic semiconductor design. Engineers and materials scientists shrank electronic components, primarily manufactured from silicon, germanium, and gallium arsenide, and combined them with light-emitting diodes (LEDs) that could be assembled into pocket-sized computers. Progress in optoelectronic technologies, which combine optical components, such as lenses and lasers, with electronic ones, and organic semiconductors, which contain, among other elements, at least carbon and hydrogen, is enabling new generations of smart devices. Some of these devices may introduce new capabilities, and others may mimic biological tissues or interact with them. Stephen Forrest, at the University of Michigan, has been studying organic and inorganic optoelectronic materials since the late 1970s. A professor of electrical engineering, physics, and materials science, Forrest directs the university’s optoelectronic components and materials laboratory. He began his career at Bell Labs, where he designed photodetectors for optical communications before moving into integrated optoelectronic circuits and organic semiconductors. This latter category includes next-generation solar cells and organic LEDs, which, unlike traditional LEDs, can be deposited onto flexible plastic substrates and have made their way into foldable smartphones and rollable televisions. PNAS recently spoke to Forrest, who was elected to the National Academy of Sciences in 2016, about his current research.

Stephen Forrest. Image courtesy of Michigan Photography, University of Michigan.

PNAS: Your Inaugural Article (1) describes a process for transforming a 2D inorganic photodetector array into three dimensions on a curved surface that mimics the eye. What inspired you to manipulate circuits in this way?

Forrest: This work speaks to my background in organic materials, because they allow for flexible and conformable devices. So it was natural for me to think about opportunities with inorganic materials as well. We don’t generally attach inorganics to oddly shaped objects because of their nonconformability. But if you want to put a device or widget onto a human being, for example, you must address these issues, because people don’t have flat surfaces and right angles.

Our group has been imagining systems that have the shape, size, form factor, and function of the human eye for about a decade. The eye is a beautiful example of a system that needs to be on a nonflat surface. The eye is small, about a centimeter in diameter, and it’s lightweight. It has an enormous capability for seeing across a wide dynamic range of light intensities with very large peripheral vision. And it has a relatively simple lensing system. These properties all come from the fact that it’s engineered to be in this shape, and that has really inspired our work.

However, imagers today are generally fabricated on flat wafers because that’s what inorganic semiconductor processing demands. We thought, “Wouldn’t it be better to make something as functional and in the same form as the human eye as an imager, instead?”

PNAS: As you develop an inorganic eye, what are some of the challenges of working with optoelectronic materials? How were you able to overcome them?

Forrest: Inorganic semiconductors are brittle. When flexed, they can suffer from deformation strain that may lead to structural damage. In addition, the increased distance between points when deforming a 2D focal plane array into a 3D hemisphere means that separation of the individual photodetectors during deformation can lead to a loss of image resolution.

This is especially inconvenient when designing an inorganic hemispherical image array, because the point of maximum deformation strain in the eye is approximately at the location of the fovea of the retina, near the optic nerve attachment. It is also the point of highest resolution. If you want to start flat and go to a nondevelopable surface (i.e., a 3D topologically transformed curved surface), the pixels that are at the foveal point undergo the maximum amount of strain, which means they are pulled further apart than any other pixels on the sphere. So you’ve lost the maximum amount of resolution right where you need it the most.

We started constructing our eye with organic photodetectors, because they are naturally flexible. What we first solved was a way to avoid this strain by allowing the pixels to stay in place while the substrate changed shape. They just slid along the surface of the hemisphere. As we thought about that result, we said, “Why can’t we do that with inorganic devices as well?”

Our solution was to introduce a topological change on the surface. For example, if you construct a circuit on a sheet of paper and then roll the sheet into a cylinder, there’s nothing topologically different between those two surfaces. Two dots on the surface of the sheet will always have the same distance between them when you roll it, if you stay on the surface of the paper.

We applied the same logic, starting with a sheet and turning it into a hemisphere. The optoelectronic components rest on top of the surface on rigid lines. Then, as the surface changes shape from two dimensions to three dimensions, those rigid lines fill in the space as the photodetectors slip along the surface. Even the highest resolution commercial complementary metal-oxide semiconductor imagers have spaces between the pixels. We just keep those spaces constant and precompensate in the 2D structure for where we want the location of the photodetectors in the final structure.

PNAS: How has working with both inorganic and organic semiconductors influenced your research group?

Forrest: I think one of the keys to the success of my group is that we are involved in both technologies. Organics give my inorganic group members inspiration and imagination, and inorganics give my organics group members discipline—how to make something that has significance and isn’t just a toy.

The constraints that you face in inorganics make you very focused on the problem at hand. Organics, on the other hand, provide an endless vista of promise and possibility. So the imagination works overtime, and some of the ideas transfer quite well across these material boundaries. We would have never imagined starting an artificial eye with inorganics because we knew the challenges were too great, but we have benefitted tremendously by keeping strong activity in both areas.