Measuring intermetallic compound

Subscribe to Thread

Hi, I am doing a research on the solder ball joint strength... - Aug 24, 2005 by

SEM to determine thickness and spectrum analyses to determin... - Aug 24, 2005 by Patrick Bruneel

Optical microscope or SEM work fine. You say, "I have f... - Aug 24, 2005 by davef

Hi, Whenever I'm measuring IMC using the optical microsco... - Aug 25, 2005 by

Khoo Meng Tze, Freescale Malaysia

#36274

Measuring intermetallic compound | 24 August, 2005

Hi, I am doing a research on the solder ball joint strength of lead free PBGA package. Coming across a dozen of literatures and journals(and the industry itself), I have found out that the method of measuring intermetallic compound(IMC) formation is quite questionable.

The current method of measuring IMC is by using an image analyzer(software) and optical microscope or SEM. Is there any other way to measure the thickness of the IMC accurately? Thanks in advance.

This message was posted via the Electronics Forum @

reply »

Patrick Bruneel

#36281

Measuring intermetallic compound | 24 August, 2005

SEM to determine thickness and spectrum analyses to determine the metals forming the IMC. This will give you everything you would want to know.

reply »

davef

#36298

Measuring intermetallic compound | 24 August, 2005

Optical microscope or SEM work fine.

You say, "I have found out that the method of measuring intermetallic compound(IMC) formation is quite questionable." What's the issue?

reply »

Khoo Meng Tze, Freescale Malaysia

#36302

Measuring intermetallic compound | 25 August, 2005

Hi,

Whenever I'm measuring IMC using the optical microscope, I have to make sure that the images captured are clear. If not, then i will have an issue in differentiating the blurry areas between the IMC and solder(for example).

This gives variation in the measured thickness and area of the IMC. Of course, if I use SEM I will get a clearer picture, but then i will be focusing on a smaller area.

I am looking at new ways to measure intermetallics or to verify the accuracy of the existing method. When you do conversion from the real physical area to pixels, there will always be some deviation from the actual. Of course the current method (using optical microscope and SEM) has been refine and I don't think the deviation is large.

Anyway, thanks for the comment.

This message was posted via the Electronics Forum @

reply »