LG 09/21/2004

Results of thinning and dicing MIMOSA wafer # B35739-09-G6

We attempted to first thin MIMOSA wafer # B35739-09-G6 to a 50 micron thickness and then dice the wafer to give us usable detectors and chip blanks that could be used to test mounting and wire bonding mechanical techniques.

The wafer did not survive the thinning process at Aptek Inc. and broke into 2 larger and numerous smaller fragments. This was attributed to the individual silicon blank quality as well as the assertion that the wafer had been pre-ground before Aptek received it. The breakage occurred near the end of the thinning process and the pieces are (as measured) ~0.0022” thick (~ 56 micron).

The wafer fragments were diced as best as was possible at Corwil Technology Corp. and the results follow.

The wafer to chip number layout is shown below;

The colors are said to designate functionality of the MIMOSA detectors as tested by probing.

Red = bad

Yellow = >/= 1 working sector

Green = all good

The detectors that survived the operation are shown below.

Detector # color condition

4 Yellow OK

9 Yellow surface scratches on sensor area

14 Red OK

15 Yellow OK

21 Red surface scratches on RDO area

26 Red bad scratches

30 Red OK

? ? OK (position on wafer not known)

This is a sufficient amount to get us started.

The diced MIMOSA all exhibit some curvature concave in the direction of the lithography. The total curvature is approximately 1 mm over the detector.

The other MIMOSA wafer (# B35739 L2 B6) will be diced prior to thinning in order to limit our exposure. The purchase order for this has already been placed.