hi Holger,

   I have recently successfully produced a homogeneous (roughness
<2nm) layer of 25nm resist by spin-coating 1:25 AZ9260:PGMEA at
2500rpm for 30sec, and then leave it for a while until the solvent
totally evaporate into air(clear and rapid change of color can be
observed), and then bake at 110 for 3min to ensure there is little
solvent or bubble left. I am quite confident that homogeneous films
down to 5nm can be produced(with a ratio of 1:50).

   The essential reason why my previous experiments failed to produce
such a good film is:

  1. <0.2um filter is indeed necessary; infact if I can find smaller
filters it will be even better.
  2. the substrate has to be extremely clean since at thicknesses
under 100nm any dust remains will cause the whole substrate to be
unusable.
  3. BARC components interact with the photoresist layer causing
undesirable chemical modifications; so they can't be used if you want
really good film.
  4. bubbles can't be avoided but can be reduced greatly by heating in
oven at a gentle temperature.

  The reason for me to find such a thin film is that I hope to do AFM
lithography on it and then do metal depositon and patterning.
Unfortunately I found that for such thin films metal lift-off will be
extremely difficult (they simply stay). I think this may be due to
pinholes. But strangely I didn't observed any pinholes under AFM
scanning. I am now looking for a solution of this.


  cheers

  Jiang


On Wed, 01 Dec 2004 00:29:50 +0100, tmm  wrote:
> Hi,
>
> as the others reported, the correct solvent is PGMEA not the AZ edge
> bead removal .
>
> But with this resist you will never be able to dilute it down to 50nm,
> no way !
>
> This resist is intended for thick layers, so exactly the opposite of
> what you want.
>
> I would try to start with the thinnest resist possible. I don´t know
> what product clariant offers for thin layers, because we also use only
> the thicker resists like 5214, 701mir, 9260 etc. from them
>
> The default thickness should not be above 600nm IMO to have a chance to
> dilute it down to 50nm.
>
> From Shipley/Rohm & Haas there are resists like UV6m UV210, UV135 etc.
> They start with lower default thicknesses.
>
> I could dilute UV6 ( 600nm standard@3000rpm) down to ~100nm, although
> the layer wasn´t that good afterwards. The practical limit was somewhere
> in the 150-180nm range.
>
> But there are some other and newer products. Check Siber resist system (
> ~150nm standard), UV210 ( 300nm standard) etc. and dilute it with the
> right solvent ( here it is not PGMEA, but DEC11 in case of UV6).
>
> But 50nm is way thin ! You will always have problems with thickness
> homogenity and defects, but it it may not be impossible.
>
> Another way...much better I think...would be to use a 2nd layer like
> Barc. Etch it in O2 Plasma and strip the resist in Aceton, but the barc
> will stay. You can easily buy Barcs in the range of 60nm and the layers
> are very good.
> If you want a sample...no problem.
>
> Why do you indend such a small layer...by the way   :-)
>
> Greets
>
> Holger
>
> _______________________________________________
> [email protected] mailing list: to unsubscribe or change your list
> options, visit http://mail.mems-exchange.org/mailman/listinfo/mems-talk
> Hosted by the MEMS Exchange, providers of MEMS processing services.
> Visit us at http://www.memsnet.org/
>