CMP Challenges

Chien-Yu Chen chienyuc at
Mon Jun 27 20:27:17 PDT 2011

Hi Ed,

I have no idea about how the membrane issue will affect the results, but I can share my recent polish rate of Epi2 Ge and LTO300PC at clean metal head sets: planar Ge is about 25-30nm/min and LTO is about 70nm/min, which are both pretty much the same as Filip tested long time ago.  If any noticeable membrane difference occurs only in gold/nonmetal heads we may conclude there is some problem, although I would prefer the pad grooves dominate the polishing rate.

Hope that helps,
----- Original Message -----
From: "Ed Myers" <edmyers at>
To: cmp at
Sent: 2011年6月27日 星期一 19:34:52
Subject: Fwd: CMP Challenges


I want to be more clear about what is going on with the CMP.  Right 
now the gold contaminated hardware set is not working.  This was 
reported by a lab member who found a zero polishing rate for this 
hardware set.  It has also been reported that the clean non-metal 
also has a low polishing rate.

I believe these problems related to the membrane assembly.  Recently 
I purchase membranes from the US distributor.  These membranes do not 
work with our hardware.  Apparently we have an older version of 
hardware, which caused the confusion.  I am working with the vendor 
to get the correct membranes, but I'm afraid they will have to be 
shipped from overseas.

In the mean time, I am also trying to address the polish rate 
variations between new and used membrane assemblies.


>Date: Tue, 21 Jun 2011 16:22:20 -0700
>To: cmp at
>From: Ed Myers <edmyers at>
>Subject: CMP Challenges
>Over the years I would get occasional reports of polishing rate 
>asperities.  This may happen more than I'm aware, since often times 
>problems or concerns go unreported.  During these isolated 
>excursions, I was never able to find any correlate to these problems.
>Recently I have changed out a number of our membranes within a sort 
>time period.  This has resulted in many more reports of polishing 
>problems.  I feel pretty confident there is a correlation to our 
>membrane assemblies and our polishing rate excursions.  I am working 
>with the US parts supplier and GnP to understand why we are so 
>sensitive to the changing of the membranes.  We are trying to 
>determine if we received incorrect assemblies, if there is a way to 
>condition the assemblies or if we need to modernize our wafer heads.
>I will keep you appraised to any findings or developments as they 
>become clear.

More information about the cmp mailing list