Well, now that you mention it, there is one ingredient in consumer
dishwasher detergent that caused problems in the past.
Sodium Dichloro-S-Triazinetrione dihydrate:
<
https://en.wikipedia.org/wiki/Sodium_dichloroisocyanurate>
It's a chlorine based disinfectant and bug killer. That's quite
useful in a dishwasher full of rotting food residue, but not terribly
helpful when cleaning PCB's. The problem was that even in such low
concentrations (about 1%), it makes a dandy bleach and was fairly good
at removing the markings off components and the rubber inspection
stamps off the PCB's. It also seemed to cause the Chemask peelable
solder mask to prematurely fall off the PCB. I could find no
conductivity or residue issues, just some odd irritations. I found a
totally functional, but highly irritating method of neutralizing the
compound. I dumped a weak acid (acetic acid or white vinegar) into
the dishwasher mixer. That got rid of the disappearing markings and
having to fish out the Chemask from the bottom of the dishwasher, but
also filled the production area with toxic chlorine gas covered by the
smell of wet athletic socks. Production declared my solution
unacceptable and found a supplier that did not use chlorine compounds
in their soap mix.
> In chemistry it was not at all unusual to refine your chemicals before
>using them for research, but that greatly depended on the research and
>the documentation detail available for the impurities. That was also
>not trusted much and verified.
I assume much of the same is done in industrial and consumer products.
However, I have no experience with producing the original compounds
and don't know how much care is taken in removing or even identifying
contaminants.
>Any idea what contaminants would matter here? Sure seems like H2O is
>the one of most concern. Heck, even the water used would have numerous
>contaminants.
Well, the dishwasher runs on water, so I doubt that's considered a
problem. Some of the junk found in the water is potentially a problem
especially since I considered Santa Clara CA tap water undrinkable and
unsuitable for decent coffee or tea.
We had a high flow water filter before the dishwasher that was good
for removing boulder size contaminants, but did nothing for dissolved
chemicals and pollutants. I also didn't have a proper lab for
analyzing what was in the water.
I did manage to create a big problem when I let the dish washer drain
the hot water heater, which pumped all the calcium carbonate from the
bottom of the tank into the dishwasher. The PCB's looked like they
were covered with white paint. I managed to recover most of the
boards with a smelly white vinegar rinse, to produce calcium acetate,
carbon dioxide, and more water. By itself, calcium acetate is fairly
harmless and somewhat soluble in water. So, I tried giving the boards
a rinse in an organic solvent. Big mistake. Mixing alcohol and
calcium acetate produces a white, slimy, sticky, and smelly gel, that
refused to be removed by anything less than explosives. Ok, lesson
learned.
Anyway, if you're trying to get down to megaohms per square sheet
resistivity, things like contaminants are important. However, if your
circuit can tolerate less, the contaminants are much less of a
problem. I've been fortunately and only had to deal with fair low
impedance designs that don't have such problems. However, if it is a
problem, you're probably better off etching a guard ring around op amp
inputs than trying to permanently clean the PCB soldering process.
Marginally related but interesting stuff:
Design femtoampere circuits with low leakage, part one
<
http://www.edn.com/design/analog/4368681/2/Design-femtoampere-circuits-with-low-leakage-part-one>
INVESTIGATION OF FACTORS THAT INFLUENCE CREEP CORROSION ON PRINTED
CIRCUIT BOARDS
<
http://thor.inemi.org/webdownload/Pres/SMTA_Pan_Pacific-2012/Creep_Corr_paper.pdf>
"The present work has shown that the presence of organic
acid flux residue is the single biggest contributor to
copper creep corrosion".