Page 20 - ChipScale_Jan-Feb_2020-digital
P. 20

Table 1: Summary of dX, dY error statistics for R1,
        R2, R3.
        Die placement error measurement
          With the accuracy of the measurements   Figure 5: Die placement error (mm) a) dX, and b) dY heat maps.
        and the correspondence between the   processing.  T he  disr uptive  EMC   Data analysis, yield prediction and
        exposure and measurement coordinate   curing effect renders the ty pical
        systems established, it is possible to   g lob a l a l ig n m e nt s olut io n u s e d   throughput
        evaluate the placement errors of die on   by lithography tools useless, as  it   The process specifies overlay errors less
        FOPLP panels. Dummy die embedded   assumes linear corrections for scaling,   than +/-15µm. As long as the measured
        in molding compound to form a FOPLP   rotation,  orthogonality.  Only  site-  errors are within the correction capability
        substrate  were supplied courtesy of   by-site corrections can accommodate   of the stepper, it is possible to yield 100%
        ESWIN. Measurements of two points   these types of errors. To apply site-by-  by correcting each die, so called “die-
        for every die, top and bottom, allowed   site corrections, each exposure needs   by-die” exposure individually, but this
        the system to calculate translation   to be matched to the local errors to   imposes an unacceptable penalty on
        (X,Y) and rotation (ϴ) offsets for each.   provide good overlay.      stepper throughput. Increasing the size
        The software algorithm generates                                      of the exposure to cover multiple die will




















        Figure 6: a) dX (mm), b) dY (mm), and c) rotation (rad) data histograms.

        “heat maps” that visualize the dummy
        measurements before cor rections
        are applied  (Figure  5).  Histograms
        of the X and Y placement errors and
        rotational errors (Figure 6) show the
        distributions of these measurements.
          The multimodal distributions of
        the dX and dY histograms ref lect
        the nonlinear die placement errors
        obser ved in the heat maps. T his
        nonlinear error can be attributed to
        the epoxy molding compound (EMC)
        curing process, which is typical for
        most  FOPLP  reconstit uted  panel   Figure 7: Predicted a) dX and b) dY overlay data histogram (mm) for 3 X 3 field, 100% yield.

        18   Chip Scale Review   January  •  February  •  2020   [ChipScaleReview.com]
        18
   15   16   17   18   19   20   21   22   23   24   25