Dear Chao,

I have also faced the same things in my experiments. I have found that
current density distribution in more localized at the corners, i.e. near
sidewalls, thus metal deposition rate is also higher. You can control this
non uniformity by changing the pulse type. Try pulse reverse plating. keep
2-3 times higher reverse current. In reverse cycle, more metal will be
deplated from the corners than at the center. Although its a bit difficult
to predict the exact amount of current values, yet 10:30 mA/cm2 (fwd:rev)
current density will be a good starting value.

Also its very defiicult to plating exact 75 um, i have used CMP process to
remove excesss metal. If you can, then try polishing.

Thanks,
Pradeep

On 3/6/06, [email protected]  wrote:
>
> Dear all,
>
> I have some questions about copper electroplating. I hope somebody with
> similar
> experience can help me.
>
> I have some via holes in my su-8 layer (with bottom gold layer as seed)
> and I
> need to do copper electroplating to fill these holes. The sizes of these
> holes
> are 30um-100um in diameter and 75um deep. After plating, I always find
>
> 1)deposition rate in larger via holes is faster, which means larger via
> holes
> are always over plated when smaller ones are just about done.
>
> 2)inside each via hole, the deposition on the side wall is easier(or
> faster?) so
> that when plating is done, the top of the copper post is not flat.
> Instead, it's
> more lile a volcano.
>
> 3) it's hard to control the plating thickness - in my case, I need exactly
> 75um
> thick copper.
>
> Anybody has any idea what I can do to solve these problems? How I can get
> uniform and flat plating (with good thickness control) through the whole
> sample?