Page 32 - ChipScale_Jan-Feb_2020-digital

each design repeated 16 times. Test dies with 64 cavities each of design B were
were manually placed in the total of 64 manufactured, where the test-dies were
cavities and their position within the mounted and their alignments measured.
cavity measured. The standard deviation No significant deviation was detected
of the die position for each design is between the substrates. The positional
shown in Figure 6. error of all dies in the X direction is
The design B (Figure 4) with two shown in Figure 7, while Figure 8 shows
centrally acting spring elements showed the positional error in the Y direction.
the smallest positioning error and was, Conclusions on LIDE-generated
therefore, chosen to be used for further passive die alignment features. LIDE
investigations. A series of five substrates technology can be used to produce

Figure 6: Standard deviation of the measurement
of the positions of the alignment marks in the test
dies in relation to the ones in the substrate wafer,
for each die cavity design with integrated passive
alignment structures.
defect-free passive die alignment
structures at the edge of open cavities
i n gla ss mou nt i ng wafe r s. Eve n
though this was solely an early-
stage study, LIDE’s on-the-fly laser
process enabled the production of each
cavity in approximately two seconds,
making it already a high-throughput
technology. Cycle times are, therefore,

Figure 7: Die positional error in the X direction.

Figure 8: Die positional error in the Y direction.
expected to be greatly improved with
a completely optimized process.
Theoretical derivations and practical
tests have shown that if the spring
elements are suitably designed, they
remain intact while in operation. In the

30
30 Chip Scale Review January • February • 2020 [ChipScaleReview.com]