Homebrew PCBs

Wed, 19 Apr 2006 19:48, lcdpublishing wrote:
><snip Tarn-X>
>I have a hunch there are two key elements to this whole ink-jet
>printing thing..
>
>1) The proper type of ink - solvent based as opposed to water based

But water _is_ a solvent, and a darned good one at that :-)

However, it's a highly polar solvent, in contrast to the non-polar
hydrocarbons and the non- or slightly-polar other organics.

>and pigmented as opposed to dyes.

But pigments aren't dissolved (it's what distinguishes them from dyes.)

So both your criteria are rather confusing.

>2) Curing that ink so that it doesn't wash away.

Agreed, but you are relying on the binder in the mix: the "solvent"
is actually a carrier liquid, the pigment is the colour (which may
be irrelevant to the resist properties) and the binder is the glue
residue which holds the pigment particles onto the surface -
or the pigment would be as durable as a dry-wipe marker
on a smooth surface.

Because these inks are designed for a high-gloss finish, there
is a higher amount of binder as a lacquer to glaze the surface,
otherwise the printing appears dull against a gloss paper.
That's probably why we are having some success with this
family of inks as etch resist.

It may also explain why yellow is best - it is a very transparent
colour which may contain little or no pigment, whereas black
requires much. It's even possible that a heavy pigment load
weakens or perforates the binder layer.

Another approach would be to reinforce the ink: copperplate
print etchers of old used to draw with a grease pencil then
dust the image with "dragon's blood" (a resin powder) and
fuse, repeating as necessary to produce a nitric acid resist.
I wouldn't want to exercise that amount of skill but it might
be made to work, perhaps even with toner powder. The old
Kodak Diconix ink was very oily/sticky but the technology
is likely too old to resurrect.

A method AFAIK untried would be to coat the board with a raw
resist coating such as gelatin, albumen, gum arabic or PVA
and then jet-print with a tanning agent. Potassium dichromate
followed by UV harden, alkaline develop and bake would
probably work but I don't like to think what it might do to the
print head. Tannic acid might work. Boiled oak-galls might.
Eye of newt, etc. - wish we had a good chemist on the list. :-)
It might be worth asking on sci.chem

>The little bit of playing around I did with the durabrite ink
>with "Bone stock" PCB material showed me that it not only adheres,
>but it doesn't bead up or puddle like the other inks.

Beading and puddling is a surface energy phenomenon, an effect
mediated by the "wettability" of the surface by the liquid and
the surface tension of the liquid, both being the manifestation
of  intermolecular attractions.

So both surface preparation and liquid composition play a part.

In respect of an aqueous liquid, the wettability of copper is
decreased by oily contamination of the surface and
increased by the presence of somewhat polar organics
such as alcohols or detergents as surfactants.

A "clean" copper surface is probably oxidized or carries
some other copper compound as a chemical tarnish derived
from its last bath - the wettability of this will be influenced
by chemical reactions as well as surface affinity.

So weak organic acids such as citric or acetic will probably
increase wettability in many situations and ammonia water,
which is alkaline and complexes copper, in others.
Acetic acid and ammonia seem to have solvent activity too.

Another possibility is a mild reducing agent

(It's subtle influences like this, combined with pre-treatment
of the surface, that makes litho printing possible.)

>So, I wouldn't worry too much about the Tarnex at this time.

I hold no brief for Tarn-X but it is worth considering what
pre-treatment might leave a molecular film to improve
the printing - it's easier than modifying the ink, because
it doesn't affect the ink delivery. Even something as
simple as buffing with a glycerin or hard-soap impregnated
cloth might have an effect (I only say that as an example
of technique, I don't know whether it would be beneficial.)
Microfibre cloth seems a good candidate: no lint and
a very fine fabric structure to dispense the agent.

This reasoning also leads on to the conclusion that when
a successful technique is found it will be necessary
to copy it with scrupulous attention to detail.

Regards, LenW
-- 
   Wisdom, free to those with the capacity to receive it:
   http://www.google.co.uk/search?&q=etiquette+%22top+posting%22