LiNbO3 is pyroelectric, so it will become charged due to temperature
transients. A slow temperature ramp is certainly possible. but that would
require a very long bake times (several hours), and the resist would end up
being over-baked. More importantly, even under ramp conditions, there are
reasons to believe crystal defects are still induced, so a room temperature
lithography is what is ideally desired.

The HMDS vapor was applied at room temperature. There as no heat. The vapor
pressure of HMDS at room temperature is about 7 Torr, so it will easily
easily boil under vacuum. Heat certainly helps to dehydrate the surface, but
I don't believe it is necessary for the HMDS to vaporize.

The theory of internal diffusion of solvents is quite possible. I might be
drying the outer layers while the lower regions (the ones that matter for
adhesion) may still have a high solvent content.



On Fri, Aug 14, 2009 at 7:30 AM, Edward Sebesta  wrote:

> Some questions.
>
> 1. When you do the vapor prime, do you use heat? You need to have heat
> so you don't leave a film of excess HMDS molecules. There are two
> theories out there how HMDS works, but you need to bake the wafer for
> either theory.
>
> 2. In the vacuum bake you run a risk of the surface of the resist
> "crusting." That is you desolvate the top layer and that film blocks the
> further loss of solvent from the bulk of the film.
>
> 3. Finally without heat, you will have low internal diffusion of
> solvents from the bulk and I would think that your vacuum baking might
> have to go for a long time, too long for a practical process.
>
> 4. I don't understand why you think softbaking would cause a charge
> build up. A hot plate shouldn't build up charge and a largely
> atmospheric oven bake shouldn't build up charge.
>
>
> Ed