Floating Method Enables Flexible Displays

Alien Technology (Hayward, Calif.) is offering fluidic self-assembly (FSA) as a method for assembling displays and other electronic systems on a variety of substrate materials, including flexible plastic sheets. Fabrication of the semiconductor components is done using normal wafer fabrication processing, instead of thin-film processes on glass plates. Individual blocks, which Alien calls nanoblocks, are then separated from the wafer. These blocks are then "floated" in a solution, and literally poured over a substrate with appropriately shaped holes for the nanoblocks. Thin film metalization is then performed to connect the blocks, completing the system. The technology originally was developed at the University of California at Berkeley, and transferred to Alien.

First, the nanoblocks are fabricated on the wafer, using standard CMOS processing. Then the wafer is thinned using standard grinding and etching techniques. A mask is then applied to the wafer backside to define the streets, and a potassium hydroxide (KOH) etch is performed, giving the nanoblocks bevels along crystal planes, as shown in the Figure. The blocks are tested on the wafer, so only known-good blocks go on to assembly.

The finished nanoblocks are then placed in an aqueous solution with additives that keep them from sticking to each other or damaging each other. Additives to promote adhesion between the nanoblocks and the intended substrate are included in the solution. The substrate can be just about any material, and the holes can be made in any way, as long as they are the appropriate shape for the nanoblocks. Standard thin-film metalization techniques, such as aluminum sputtering and etching, can be used to connect the blocks once they are in place. Nanoblocks that do not assemble are simply cleaned and put back into the process.

Since the nanoblocks are small, on the order of 100 µm, and thin, they are very robust and can withstand bending. They can be assembled in a flexible substrate, and the substrate can be bent without worrying about ruining the display. These displays can be installed on smart cards, flexible membranes, paper-like or roll-up sheets, or any reasonably smooth surface of any shape. A wall-hung roll-up display would be possible.

Most displays are transmissive; they are backlit, and the pixel elements block that light when activated. This assembly method lends itself well to reflective and emissive display technologies.

The nanoblocks are pad-limited, since the interconnect pitch between the blocks is so much larger than the on-chip technology. This allows the nanoblocks to contain the extra logic required to be rotationally symmetric in a functional as well as physical sense. Since the beveling goes along crystal planes, the blocks are restricted to square and rectangular shapes. Two-fold rotational symmetry (180°) is the absolute minimum, and four-fold symmetry (90°) is required for the square nanoblocks, which are likely to assemble more reliably in this process.

The process is compatible with inexpensive production methods. The company uses a roll-to-roll plastic sheet process, which is much less expensive than typical glass panel processes.

This technology is not limited to display applications. Non-display systems could be made on a sheet. The company has proposed this method for assembling III-V laser devices onto silicon substrates, which would contain the driver circuitry. Many possibilities exist for this low-cost assembly technology. •