Hi Sebastian
I doubt that there is be a problem with your AlN layer. Especially if
you are checking it using XRD.

AZ400k is known to etch AlN as the active etchant in it is KOH (KOH etches AlN).
The following reference might help you:
"Wet chemical etching of AlN"
JR Mileham, SJ Pearton, CR Abernathy, JD MacKenzie … - Applied Physics
Letters, 1995

Hope this helps
Nipun Sinha

On 8/6/07, sebastian wicklein  wrote:
> Colleagues,
>
> I'm manufacturing micro-capacitors. In doing so I put down an AlN layer as a
> dielectric between the plates via reactive sputtering.
>
> During manufacturing I need to pattern the AlN-surface with Photoresist
> AZ9260.
>
> The problem I'm having at this step, is, that my Developer AZ400K 1:4
> (Developer : H2O) @ 27C takes off the AlN layer for some reason. It attacks
> it and makes it peel off.
>
> However, XRD shows a clear 0002-peak and EDX is fine. The layer is 2microns
> thick. Underneath the AlN you can find Cu, or Al as a capacitor plate or
> highly B-doped Si.
>
> I need to mention here that I don't have the possibility to take off the
> native oxide with a plasma prior to AlN deposition.
>
>
> So, what is the problem with my AlN layer and how can I improve my layer?
>
> Might the native oxide on the above mentioned materials cause that effect?

##################################################
Nipun Sinha
Graduate Student
Mechanical Engineering and Applied Mechanics
University of Pennsylvania, Philadelphia, USA
##################################################