Natsuki,

   There are a couple of things you should consider doing to keep the
surface clean.  First, coat the surface with a thick layer of resist,
say 5um, before dicing.  Then solvate the resist to remove surface
debris.  Also, I think you will have much better luck at shorter
wavelengths 248 or 157nm where more of the energy goes into direct bond
breaking instead of just heating the sample which results in thermal
damage at the edges.  I would recommend talking to someone who does this
sort of thing regularly.  You might consider JPSA lasers, I've had good
luck with them in the past, jpsalaser.com

-Cheers,
    Michael



>>> Vijay Rajaraman - EWI  11/3/2009 11:39 AM
>>>
Hi Natsuki,

One trick you can try is to perform edge grinding of your laser cut
wafers.

Regards,
Vijay
(TU Delft, NL)