Origami technique turns flat optical sensors into hemispherical eyes
While vertebrates evolved with globe-shaped eyes, with a concave retina behind a spherical lens (and the iris as a pin-hole), insects mostly sport convex multi-faceted compound eyes. Both approaches offer a wider field of view and lower aberrations than our made-made planar sensors. Yet, conforming high-resolution semiconductor-based sensors into hemispherical domes (inward or outward) presents its own challenges.
Now, a team of researchers from the University of Wisconsin Madison leveraged Japanese paper folding technique “origami” to shape silicon optoelectronic sensors into near perfect hemispherical sensor arrays.
icosahedron based on silicon nano-membranes is
pressed into hemispherical concave or convex molds.
Publishing their results in Nature Communications under the title “Origami silicon optoelectronics for hemispherical electronic eye systems”, the researchers explain how they took advantage of traditional planar fabrication techniques to design an array of silicon-based lateral P–i–N photodiodes, laid out to form a large net of pentagon- and hexagon-shaped single photodetectors forming the flattened subdivisions of a half-truncated icosahedron (think of a soccer ball cut along carefully chosen seams).
Once lifted off their original substrate and transfer-printed to a flexible one, the thin-film array of photodiodes was cut-out along specific contours using a precision laser so the subdivisions’ edges could then be jointly folded into a perfectly matching concave or convex hemispherical mould.
Using this approach, the researchers fabricated single-crystalline silicon-based focal plane arrays (FPA) and lens-less artificial compound eyes (both with hemisphere-like structures, but inverted).
bends inward for creating a hemispherical focal plane
array while the convex version (right) mimics an insect’s
compound eye. Yei Hwan Jung and Kan Zhang.
Interestingly, because the active layer are so thin, the same folding mechanism can be implemented for both concave and convex photodetector arrays. The paper also highlights that the origami-based fabrication eliminates the use of metal wires in-between pixels, for the connection of sparsely arrayed devices, instead the jointing folds allow for densely packed pixel arrays. For a better fit, as design resolution increases, the edges of the hemisphere-like structure can be further smoothed out by dividing large pentagonal and hexagonal faces into smaller polygon faces.
photodiodes, with a metal connector waving from
one cell to the next. Scale bar in the microscope
image is 50μm.
In their implementation, the researchers packed 281 hexagonally-shaped photodetectors (each 113 μm in diagonal) in hemispherical arrays of different radii, 2.27mm and 7.20mm, only observing a performance degradation when the photodetector was conformed to a curvature radius of 1.5mm. But they note that if made thinner (down to 20nm), a silicon nanomembrane photodetector could easily wrap around a single mode fibre (125μm in diameter).
origami sensor draped over a dome shape, with
perfectly jointing seams. Yei Hwan Jung and Kan Zhang.
The hemispherical FPA was assembled into a simple camera system featuring a plano-convex lens (10mm diameter and 10mm focal length) while the convex version was evaluated as a compound eye mimicking camera (a photoresist microlens placed on top of each detecting unit during fabrication to maximize light intake).
The paper highlights the simplicity of the origami-shaping process and how easily the pixel density could be scaled up. The researchers hope that further optical optimizations of their compound electronic eye system (such as adding layers to mimic the pigment cells and crystalline cones) could yield truly panoramic colour vision.
College of Engineering University of Wisconsin Madison – www.engr.wisc.edu
Related articles:
Sony curves images sensors; TSMC stacks them
Thin film micro-lenses stretch to focus
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
