photolithography or more precise alternative

[Thomes]

how can I print a pattern on a film or transparency with one micron resolution, and can I acid etch (as in printed circuit boards) to that resolution.

 

[Thomes]

it seems to me that the etching would only be constrained to the size of a photon (during exposure), and the size of an acid molecule (during etching). Of course theory is only good in theory.

 

[itsmoked]

Naw you can't etch to one micron on an edge with acid, I wouldn't think. The semi people are laying down layers like paint and that is how they can achieve the control. Or if they are etching they usually use a gas.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[sreid]

It can be done but is a specialized industry.

http://www.drc.com/metrigraphics/metrigraphics.htm

 

[IRstuff]

Not for a printed circuit board.

The copper is SUBSTANTIALLY thicker than the comparable metallization on a 1 micron design rule IC. Since the sideways etching is roughly comparable to the downward etching, with 1 oz copper being about 36 microns thick, the starting linewidth to achieve 1 micron lines would be about 73 microns.

AND, there's no guarantee that the resist won't lift altogether and break your line.

TTFN

 

[MikeHalloran]

The people who make rotary encoders and linear scales for machine tools are apparently able to get single digit micron resolution depositing chromium on glass.

So, surely, _someone_ can do it, but they might be reluctant to share their secrets because of their commercial value. Which, in turn, means that you can probably just _buy_ what you want.



Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

Yes Mike, but you said deposit not etch.

The difference is like climbing a tree or falling out of a tree.. One is real easy.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[IRstuff]

Chrome on glass is probably etched, but the thickness of the chrome is in the single digits of microns, at most.

PCBs could be etched with higher accuracy, but you'd need about $5 million worth IC processing equipment to do it:

10x or 5x photolithography stepper
PMMA photoresist system and developer
Reactive ion/sputter etcher.

The main limit is the thickness of the copper. A 1 micron thick layer would be easier to etch.

TTFN

 

[richs]

Hiya-

You will not get 1 micron out of "Photolithography" in the
standard sense. The old stuff (glass masks) were ALMOST
that good.

Temperature control, LOTS and LOTS of trial and error.

Film, no way. Edge effects of the "grain" will blow your
requirements out of the weeds. I've used some "good"
kodalith, mighty fine lenses and some really good photo
enlargers and I still get grain limits.

There are tricks that you can play with multiline, out of
phase limiting marks/lines that will allow better resolution
that the raw marks.

Also, reference against a resolution error table based
upon calibration with a laser interferometer can be used.

Still, there is a reason that this sort of accuracy costs
so much.

As an aside, you might look into a Fabry-Perot Interferometer
as an alternative.

That's pretty good at a relative displacement thing around
a couple of hundred nanometers.

Hummmm, I wonder what the track to track spacing of a
typical CD or DVD is? Look at the galvenometer reading
of a DVD player solid state laser assembly......

Hope this helps.

Cheers,

Rich S.