I'm using Argon plasma etching to increase layer to layer adhesion during Ni electroplating with photoresist as the mask. My hypothesis is the Ar etch will remove organic material, remove NiO on previous layer, create roughness on the Ni & improve contact angel to increase adhesion.
The tricky part is that i have Cu presence in the layer 1.
I use shear testing to test the adhesion & the result has been inconsistent, while the NiO removal has been consistently effective. The result has draw me these questions:
- What's the effect of Cu presence? will it be redeposited on top of the Ni?
- What's the selectivity of Ar etching between Ni & photoresist? also between Ni to Cu?
The result has shown that adhesion is better in 4" wafer compare to 6" wafer, but increasing the time in the 6" wafer has made adhesion worse, it could be because of the Cu presence. Does anyone has a better explanation?
Hi Denni,
Look up the sputter yields rates for ion bombardment on the internet. For every Ar+ ion impinging on the surface, 2.3 Cu atoms will be removed (@600eV ion energy). For Ni this is 1.5 atoms @600eV. These rates are dependent on the energy of the Ar+ ion so you need to look up up for your process conditions but the ratio between these two sputter rates stays almost the same at different energies. This means the Ni layer will be depleted with Cu during the sputter etch. The selectivity between Cu and Ni will be in the order of 2.3/1.5 .The selectivity between Ni and resist will probably be high since you will carbonize the toplayer of resist and then the etchrate of the resist will drop really fast. If the etch is short you will probably not notice it but otherwise it can lead to very difficult to remove resist layers after the sputter etch.
Depletion of the Ni layer might have an effect on the adhesion but in general a sputter etch greatly improves adhesion by roughing up the surface. So my guess is that something else is going on if you see adhesion problems at prelonged etches.
Maybe those layers are coming out more hot from the etch chamber and will reoxidize more quickly.
Thank you, Peter. It explain a lot. I minimize the etch time to 5 sec & it works. In this case, it also make the resist removal easier.
Hi Denni,
Good to hear that I could help. Succes with your further work on the electroplating.
- Badges
- Science topic