Uniquely orienting dry micro assembly by two-stage shape recognition

We demonstrate a completely dry, uniquely orienting and parallel micro assembly method on the basis of two-stage shape recognition between complementary features on parts and receptor sites. A dry environment benefits assembly of chips with either exposed movable microstructures or materials sensitive to liquid environments; uniquely orienting self-alignment based on complementary features works for any part shape; parallel assembly greatly increases throughput for mass production. In our experiments, 1 mm square dummy silicon parts were assembled with a defect rate of /spl sim/2% in 10 minutes on each of two assembly templates having respectively 397 and 720 receptor sites. This assembly technique enables either wafer level packaging of micro device chips or easy part feeding and palletizing for robotic assembly systems without constraints of part shapes or materials.