TLDR: bulk material for standard packages can be removed using a fiber laser. However, a CO2 laser will not work (lower peak power ⇒ melts silica instead of ablating it). However the laser will damage the die surface if you try to go all the way to the surface (without using a special setup like FALIT). Finally, COB epoxy appears to be workable even with a CO2 laser

General results

  • Some settings work decent
    • Possibly COB epoxy is easier to laser than package epoxy
  • Risk of melting bond wires and/or die
    • Use lower power and/or pulsed?
  • Practice on samples before trying something you care about
  • Puts off nasty fumes. Make sure to have good ventilation

From Sam Wagner:

  • “We use a yterrbium fiberlaser. What you are doing is trying to cause the plastic compound to explosively eject the silicon beads. So a pulsed laser is critical. The laser melts the plastic and the pulse hopefully causes it to vaporize and eject beads.”
  • “Important to note that if you see silicon, you’ve destroyed the chip and can now no longer do FA work. We stop at the bond wires and use acid or plasma to remove the rest. Then it’s off to SEM or TIVA/LIVA or whatever you’re next step is.”


Above: laser decap from below (Source)


  • Seemed to work okay
  • What would it look like if touched up with WFNA?

EMSL 6502 test

Above: poor results. Maybe from CO2 laser (source: photo by mcmaster from EMSL test)

Silicon Exposed

Laser Decap Pro


  • Gold backing block
  • Good leadframe exposure
  • Die unclear how well exposed
  • Scanned entire chip (including die area) up until the end
  • Raster scan is horizontal, vertical, and diagonal

Sesame 1000


  • Cavity into pic
  • Gives a close up at the end

CLC ControlLaser FALIT


  • Laser Decapsulation FALIT System Software | Image Templates
  • Pro system demo
  • Has nice features like overlaying x-ray, CSAM, or other guiding image
  • Laser scan was horizontal, vertical mix

Light Ray


  • They do the whole chip but avoid die area
  • Light Ray - IC Decapsulation

mcmaster ezlaze

TODO: add pictures, estimate material removal rate

Basically this was too slow to be practical except for very small samples. That said, it did work very well. A quicklase or other higher throughput Nd:YAG system would likely work well

mcmaster PLCC


Can be used to cut open package


Seems to also work

decap/laser.txt · Last modified: 2023/09/04 07:46 by mcmaster
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution 4.0 International
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki