Dear Hoaseng,
In my opinion, the problem is underplating which is related with the
adhesion of the photoresist to the surface. Do you use HMDS before spinning
your photoresist? If HMDS does not work, you may try to use adhesion
promoters such as the one offered by Silicon Resources. AP140 will probably
solve your problem.
Regards,


--
Kagan TOPALLI
Research Assistant (Ph.D Candidate)
MEMS-VLSI and EMT Group
Middle East Technical University
Dept. of Electrical & Electronics Eng.
TR-06531 Ankara Turkey
Phone: +90 312 210 44 09 or +90 312 210 23 40
Fax: +90 312 210 23 04
http://www.mems.eee.metu.edu.tr/
http://www.eee.metu.edu.tr/~emt/

----- Original Message -----
From: "Hosaeng Kim" 
To: 
Sent: Thursday, December 21, 2006 12:56 AM
Subject: [mems-talk] Gold electroplating


> Hi all,
>
> I am trying to electroplate gold on a seed layer
> (Ti/Au/Ti=500A/1500A/500A)
> after removing the top layer (Ti). After gold electroplating, the gold
> area
> electroplated was enlarged. For example, the target width of a gold signal
> line was 40um, but it became 49um after electroplating. Here is the my
> fabrication recipe.
>
> Fabrication process.
>
> Photolithography
>
> (a) Spin Shipley 1045 photoresist on wafer at 500 rpm for 15sec, then 4500
> rpm for 40 sec (4um).
>
> (b) Soft bake the photoresist on a hot plate for 1min. 30 sec at 105˚C.
>
> (c) Hard contact align (MA6 aligner) and expose 15sec.
>
> (d) Develop in MF351:DI water=1:5 for 1min.
>
> (e) Rinse the wafer in DI water.
>
> (f) Dry with nitrogen gun.
>
> (g) Use acetone to open small regions for the electrode contacts to the
> seed layer while plating.
>
>
>
> Au electroplating
>
> (h) Dip in HF:H2O(=1:4) solution for 4 sec to etch exposed top Ti layer.
>
> (i) Electroplating the Au to 3.5um thick in the electroplating solution
>
>
> I tried  to figure out the causes of the enlarged gold line, and I found a
> few things:
>
> (a) After developing PR, the width of signal line was about 42um
> =>2um=42um-
> 40um.
>
> (b) After removing the first Ti layer, there was undercut beneath the PR
> layer. The width of the signal line became was about 44um =>
> 2um=44um-42um.
>
> (c) After gold electroplating, the width of the signal line was 49um=>
> 5um=49um-44um.
>
> (Here, the width of signal line was measured by a microscope with a ruler
> on  an eye piece, and these values are approximate numbers)
>
> Based on this analysis, I believe that I have to modify the
> photolithography process (exposure time and developing time). However, how
> can I compensate the other problems?  Do you have any ideas?