Hello Eric,

My name is Mario Robles and I work for Tanaka Systems.  We sell Asahi Diamond
dicing blades and I will do my best to help improve your process.  I have a few
questions regarding your current process and will suggest some of my thoughts.

1. What saw do you use?
2. What blades do you use, Disco, Semitec....? Specification....S0820 or NBC-
ZH104J 27HDAA?
3. What metallization does the wafer have on either side?  Is there anything in
the streets? Are there streets on both sides?
4. What is the chipping spec you need and are currently achieving?
5. How fast are you cutting and current spindle RPM?
6. How is the wafer thinned to 6 mils? Diameter?
7. Are the wafers warped or bowed?

Depending on chipping spec it will be very difficult to get the same chipping
spec top and bottom using one pass cutting.  You can try playing with your saws
software to achieve two pass cutting that might help but still not the greatest.
A blade can help improve your results but will not be the solution.  One idea
would be to wax your wafer to another silicon wafer and cut through. This will
reduce the backside chipping quite a bit.

Can you tell me what company or University you work for? Also what sort of
product are these wafers for?  DO you cut any other wafers?

Thank you for any information given.  It will help me understand what we could
do to improve your process.

Mario Robles
Technical Sales Manager

650-966-8001 phone
650-966-1881 fax

-----Original Message-----
From:   YS See [SMTP:[email protected]]
Sent:   Thursday, April 27, 2000 8:33 PM
To:     [email protected]
Subject:        Bottom cracks after dicing

Hi,
I'm Eric - application engr, for wafer saw process. Recently I came
across a type of wafer known as the 'clipper wafer' whereby both top and
bottomside of this wafer is identical in order to enable contact to be
made via miniture clips thus eliminating the need for wire bonding .
Current process to ensure 99% yield , they have to do a first cut 30%
thru' with a porous chuck separated by a silk screen , flip the wfr over
then mounted onto blue tape then sawn thru 1 mil into the tape. This
process take almost 3hrs to complete. Attempt to do a single pass is
disasterous almost 99% backside crack die .Street width is 3mil.
thickness is 6mils . Currently we are using a soft nickel blade, 2 -6 um
diamond size,low concentration ,thickness 0.8mil blade with exposure of
20mils . Tape used Nitto 224 , bake temp . 70deg. for 30min. prior to
saw .
Wonder if you can offer any recommendation/info/article to handle thin
wafer with backside cracks issues . Any form of advise wil be greatly
appreciated
Thanks
Have a nice day.
Eric See