Thanks Michael.  Actually there is residual resist at the bottom of
the entire structure - where the wafer meets the first layer of SU-8.
Just to be clear: I spin and pattern the first layer, then spin and
pattern the second layer, and then develop both layers simultaneously.

The first mask (bottom layer) patterns 50um wide channels, and
100umx100um square pillars.  The second mask only patterns larger
pillars onto the 50um tall 100umx100um pillars.

-JP

On Sat, Nov 15, 2008 at 2:43 PM, Michael Larsson
 wrote:
>  Hi John,
>
> By blurry line, do you mean that the rim of the mould, at the depth of the
> top of the first (50um) layer, appears cloudy? Or do you mean there is some
> sort of residual overhang that forms at the interface between 1st and 2nd
> layers, follwoing development of the top layer?
>
> I presume the two layers are patterned to form different patterns, creating
> a stepped/terraced mould.
>
> Michael
>