Evaluation of Laminates in Pb-free HASL Process and Pb-free Assembly Environment

davef

September 25, 2012

Hello Mei Ming ... The amount of voiding seems surprising, especially with the ImAg boards, as you mentioned in the paper. It seems like all boards, except good ol’ PBb HASL, had voiding. How do account for that?

MKhaw

September 25, 2012

Yes, that's very puzzling for me too. The voids that I see are very uniform small voids, not the typical delamination of different sizes. I believe that the circles that I have captured are possibly not voids formed during board processing because they are too uniform, but are actually the fiber bundles instead. The diameter of the voids are therefore comparable to the fiber dimensions, about ~2-5 microns. Although these circles would typically appear as tubes from the top surface of the board, the expansion and contraction of the fiber bundles might have moved or warped slightly to enable it to be captured as circles by SAM instead. That's the only explaining that I was come out with so far.

davef

September 25, 2012

You're correct that it has to be something like SAM artifacts in the fiber bundles, because there's almost a repeating pattern in the voids.

davef

September 25, 2012

In talking about the laminates being tested, your paper said that materials A and B are high-Tg whereas C and D are mid-Tg materials. * What was the Td of these materials? * Did you see that these coefficients had any impact on the study results?

MKhaw

September 25, 2012

I'm sorry, but there was a printing error whereby A, B and C are high-Tg materials whereas D is the only mid-Tg material. Here are the Tg/Td for each respective material: A: 175/330 B: 170-180/335-345 C: 170-180/345 D: 135-145/305-312 These are typical values that I gather from their datasheets. The only obvious correlation that I can see is that mid-Tg (D) has a higher possibility of delamination compared to its high-Tg counterparts, especially after being exposed to multiple thermal cycles.