PCB without solder mask copper corrosion question

[cdev]

I have a custom device which has been in a drawer for at least five years part of which it turns out was made without any kind of conformal coating or solder mask and is now nonfunctional. It basically consists of a programmable board - a commercial board - very similar in size and function to one of the Blue Pill STM32 boards (it uses a PIC 16F877A  chip) which looks to be in pretty good shape, and the homemade carrier board, which has only its very thin traces on it which are now green. What I would like to know is what happens when thin copper traces corrode? There are a lot of traces, very thin ones. I'm actually almost afraid to touch it, they are that close to falling off. There may be open traces. The traces are so incredibly thin. A bit of IPA was able to remove some but not all of the green efflorescence. I am wondering if I should try to clean it in some distilled ('deionised') water. That has worked for me in the past with boards that did have conformal coatings. But this board looks particularly fragile and I'm afraid that any abrasive activity could easily end up removing a trace. It was working when it went into the drawer, but its currently non-working. Its a custom data logging tool I'd like to have the use of again..

If I can get it working and get most of the gunk off, my plan is to apply some solder to the traces to give them a bit more body/size and and then what? I am thinking that if it works it needs to then get solder mask to prevent any further erosion and basically, keep the traces on.

Does that sound like a good idea? Either that or try to get a replacement and maybe have to re-write some code.

[dzseki]

It is hard to tell how to fix the current board, but tin coating would help from further corrosion. There are solutions for DIY PCB makers either to have tin coat bath, or "lötlack" you can spray on and it prevents the board from corrosion, while let to solder still.

[cdev]

Googling "lötlack" it appears that its a popular brand in the German speaking world.

I have a tiny bottle of brush-on conformal coating that I have barely used because most of my boards are so small Ive just coated them thoroughly with solder and that has been enough.

But I should get some of the spray on stuff, just to have around.

I think solder mask or tinning is pretty much a necessity for any commercial product.

Also, make traces thick enough to last under typical conditions,  not as thin as can be done physically.

[T3sl4co1l]

Efflorescence?

It's screwed, man.  Somehow chlorides got into it.  It's not going to stop oxidizing.  A coating can only slow it at this point.

Tinning will help (hermetically sealed surface!), but it's underneath the traces too, and oxygen will diffuse through the laminate and soldermask.  Clean it the best you can, maybe not with an ultrasonicator, but a long soak in a few percent sodium bicarb maybe, or a solution of disodium EDTA maybe, if you can find it (something that contains it, a cleaner or polish maybe), assuming that won't accelerate corrosion of course (I'm not sure -- test on something else?).  Then wash with clean water, dry with alcohol, then bake and coat.

Better in the long run to find another, or if it's unique as far as you know, RE and copy it.

Tim

[cdev]

See picture.

Normally I have a dehumidifier running in my basement, but in the cold months the dehumidifier doesn't activate when it should, it doesn't work properly.

This was inside of a cabinet, in an only marginally heated area that gets fairly cold in the winter.

This picture is of the bottom of the board which was hidden under a sort of case it came in. If I had seen this earlier I would have been able to prevent it. The top of the board has a little bit of green stuff on it but at first glance, superficially it looks semi-okay.

Going to have to look this up. Glad I caught this now.

[T3sl4co1l]

Yeah, probably mixed chloride-hydroxide, who knows where the chloride came from.  Could just be sweaty dust, could be activated rosin, who knows.

Looks tinned with pinholes that are corroding, but metal finish is hard enough to photograph in good white light, let alone just taking a picture anywhere.

Could also be coated with a clear lacquer -- Tek used a clear soldermask in the 70s -- and again corroding at pinholes.

Soldering looks bad too, unclear if it's messy rosin or bad soldering or corrosion or what.

Tim

[tooki]

It is hard to tell how to fix the current board, but tin coating would help from further corrosion. There are solutions for DIY PCB makers either to have tin coat bath, or "lötlack" you can spray on and it prevents the board from corrosion, while let to solder still.

The Kontakt Chemie "Lötlack" is essentially a spray flux which I assume is non-hygroscopic, so probably rosin or synthetic resin. (It's not a conformal coating as such.)

You can achieve much the same effect by just applying liquid rosin flux with a brush and letting it dry thoroughly. (I've made flux of just rosin and IPA for this purpose, and it works well, but takes a while to dry fully. Might try adding some acetone and naphtha next time to speed up drying, as Kontakt does.)

[T3sl4co1l]

That doesn't sound very good...

There is OSP (Organic Surface Preservative), not so much a coating as a chemical that adsorbs onto the metal surface, preventing oxidation.  This works for bare copper boards that may be in storage up to a year (or so, I forget what exactly) before assembly.  I don't know that it's any good for long term protection.

I've had paste rosin deposits turn green, due to corrosion.  Which makes sense, because rosin is a mixture of organic acids, which eat up and dissolve oxides when soldering -- hence fluxing -- but as long as they are the least bit mobile, and oxygen is around, they will corrode.

I haven't noticed any obvious corrosion under a hard crusty rosin deposit, which makes sense.  But fresh paste tends to stay gooey (depending on what carriers and solvents with high boiling points were used, if any), and that is also the consequence I have observed.

So, neither is very good, chemically.  I think just a clear lacquer will be much better, to block the growth of corrosion spots and reduce the access of air, moisture and corrosives to the surface.  Assuming you can remove the corrosion spots, and most of what's causing them, of course.

Maybe the suggested material is very different, in which case this doesn't apply.  Dunno.

Tim

[tooki]

That doesn't sound very good...

There is OSP (Organic Surface Preservative), not so much a coating as a chemical that adsorbs onto the metal surface, preventing oxidation.  This works for bare copper boards that may be in storage up to a year (or so, I forget what exactly) before assembly.  I don't know that it's any good for long term protection.

I've had paste rosin deposits turn green, due to corrosion.  Which makes sense, because rosin is a mixture of organic acids, which eat up and dissolve oxides when soldering -- hence fluxing -- but as long as they are the least bit mobile, and oxygen is around, they will corrode.

I haven't noticed any obvious corrosion under a hard crusty rosin deposit, which makes sense.  But fresh paste tends to stay gooey (depending on what carriers and solvents with high boiling points were used, if any), and that is also the consequence I have observed.

So, neither is very good, chemically.  I think just a clear lacquer will be much better, to block the growth of corrosion spots and reduce the access of air, moisture and corrosives to the surface.  Assuming you can remove the corrosion spots, and most of what's causing them, of course.

Maybe the suggested material is very different, in which case this doesn't apply.  Dunno.

It's not the rosin itself causing problems. As you know, rosin flux residues (from rosin core solder) can stay on circuits indefinitely without causing corrosion. So can no-clean fluxes, or liquid rosin fluxes, as they dry out completely, leaving behind only rosin.

Paste fluxes, it depends on what's in them!

And for sure, in a spray designed specifically to prevent corrosion, it's fair to assume it's not going to cause corrosion.

[cdev]

There was a lot of shiny flux residue on there, which I cleaned off with IPA first thing after the device wouldn't boot.

Do you think I have a good case to ask the manufacturer to replace it? They have a much more professional looking product now.

[tpowell1830]

This is cuprous oxide, so the problem is that you only see it where it escaped from under the tin. The copper becomes very brittle in severe cases. Since it is tinned, you will not be able to clean the surfaces, although that will not fix the problem anyways. The corrosion is well inside the copper and there is no saving it now. This is also a problem when you have silver plated copper in wires.

As far as whether the company will replace it, that is anybody's guess, good luck with that.

EDIT: Conformal coating would not and will not save this, it is a result of tinning the copper, perhaps with a 2% silver solder, that created the problem. The mfg may know that this is a problem on this board.

[cdev]

2% silver solder causes this problem?

I had never heard that and indeed, I have been using eutectic alpha-fluitin solder that is 2% silver in much/most of my own stuff.

[tpowell1830]

2% silver solder causes this problem?

I had never heard that and indeed, I have been using eutectic alpha-fluitin solder that is 2% silver in much/most of my own stuff.

Well, silvered copper does not always develop cuprous oxide, in fact it is rare, but it does happen. We have a cuprous oxide program for incoming wire to check for it on silvered copper wire and cables and if it is found, the entire roll is returned.

[T3sl4co1l]

Silver plating wire is a very different process from soldering circuits...

Tim

[tpowell1830]

Silver plating wire is a very different process from soldering circuits...

Tim

Yes there is a difference, but how does that change the facts? I am interested in hearing your point.

[cdev]

Its alive!  I got it working..

I think one of the traces was no longer connecting. Almost invisible.

Now I need to figure out what to do to keep it working.. Coat it with my conformal coating / solder mask stuff, maybe. Need to figure out what to do there.

Thanks for the help!

[T3sl4co1l]

Yes there is a difference, but how does that change the facts? I am interested in hearing your point.

Corrosion is a surface process -- the chemicals, prep and methods used in wire manufacture are very different from those used in PCB manufacture and assembly.

And, surface chemistry in general, is notoriously... dumb.  Bad, hard, whatever.  It's individual molecules working away.  It's hard to study, and I'm not aware that anyone has ever had more than a reasonable idea what's going on under very, very contrived conditions.  (Which can sometimes be very interesting conditions indeed, like writing the IBM logo on a gold? crystal face.)

We can only guess what's going on here, of course -- the active sites are probably microscopic and intergranular.  It takes highly specialized equipment to tell what's going on.  I have only guesses, either way.  I would be quite surprised if, even an expert in the field of cuprous alloy corrosion, had an exact description of what's going on here, based on such basic information.  (They have likely seen something similar before, but also seen enough different situations which produce the same symptom, thus ruling out an exact conclusion.)

My guess for your example, something related to the processing of wire and silver plating.  That might involve electroplating in a silver nitrate bath or something like that.  It might involve an etch step (besides the natural displacement that occurs when copper is introduced to a silver plating bath), which would be a prime way to trap microscopic amounts of contaminants.  If the wire is worked after plating (it could be drawn to smooth or harden the plating; I have no idea if they actually do that), that could expose those contaminants through microscopic cracks, and hence compromise the strength and corrosion resistance of your product, leading to lot rejection.

So, making a direct comparison between the two products is tenuous at best.  Certainly not deserving of a direct and absolute phrasing.

My snap reaction is to dismiss --
1. Factually wrong statements (e.g., "this is cuprous oxide" when the deposit has been described as "green", not brown?);
2. Statements made without any possibility of an alternative (note my speculative phrasing, "it might involve..");
3. Specious and unsupported conclusions (e.g., " 2% silver solder, that created the problem" -- says who?  Such solders have been in use for decades (heck, probably millenia, if not intentionally and knowingly so), for various purposes, and are not generally known to accelerate corrosion over other alloys.)

And yes, I realize my statements are just as much informed by the piecemeal experience I have, as yours are.  I've presented very little of that here (I did at least support my conclusion of chloride involvement: copper oxychloride meets such a description, is easy to produce (I've made it myself), and is well understood), but if you want to call me on anything I'll be happy to find a reference.

Tim

[tpowell1830]

Yes there is a difference, but how does that change the facts? I am interested in hearing your point.

Corrosion is a surface process -- the chemicals, prep and methods used in wire manufacture are very different from those used in PCB manufacture and assembly.

And, surface chemistry in general, is notoriously... dumb.  Bad, hard, whatever.  It's individual molecules working away.  It's hard to study, and I'm not aware that anyone has ever had more than a reasonable idea what's going on under very, very contrived conditions.  (Which can sometimes be very interesting conditions indeed, like writing the IBM logo on a gold? crystal face.)

We can only guess what's going on here, of course -- the active sites are probably microscopic and intergranular.  It takes highly specialized equipment to tell what's going on.  I have only guesses, either way.  I would be quite surprised if, even an expert in the field of cuprous alloy corrosion, had an exact description of what's going on here, based on such basic information.  (They have likely seen something similar before, but also seen enough different situations which produce the same symptom, thus ruling out an exact conclusion.)

My guess for your example, something related to the processing of wire and silver plating.  That might involve electroplating in a silver nitrate bath or something like that.  It might involve an etch step (besides the natural displacement that occurs when copper is introduced to a silver plating bath), which would be a prime way to trap microscopic amounts of contaminants.  If the wire is worked after plating (it could be drawn to smooth or harden the plating; I have no idea if they actually do that), that could expose those contaminants through microscopic cracks, and hence compromise the strength and corrosion resistance of your product, leading to lot rejection.

So, making a direct comparison between the two products is tenuous at best.  Certainly not deserving of a direct and absolute phrasing.

My snap reaction is to dismiss --
1. Factually wrong statements (e.g., "this is cuprous oxide" when the deposit has been described as "green", not brown?);
2. Statements made without any possibility of an alternative (note my speculative phrasing, "it might involve..");
3. Specious and unsupported conclusions (e.g., " 2% silver solder, that created the problem" -- says who?  Such solders have been in use for decades (heck, probably millenia, if not intentionally and knowingly so), for various purposes, and are not generally known to accelerate corrosion over other alloys.)

And yes, I realize my statements are just as much informed by the piecemeal experience I have, as yours are.  I've presented very little of that here (I did at least support my conclusion of chloride involvement: copper oxychloride meets such a description, is easy to produce (I've made it myself), and is well understood), but if you want to call me on anything I'll be happy to find a reference.

Tim

I don't care to call you on anything, merely asked your point. If you believe this is copper oxychloride, I did not and do not dispute your claim, nor does it matter to me. If you want to stick to that claim, who cares? This is just a discussion on a forum, no need to get upset and start all this very extended reply. I merely stated that this is cuprous oxide and if this has offended your sensibilities then I guess you have to deal with that, but to spend 7 paragraphs doing so is really not a good discussion, but looks very defensive. I have experience with cuprous oxide and it could be that, but even if it is copper oxychloride as you say, the traces are no longer viable, and that really is the only point needed. Also, I did not say that 2% silver caused this problem as you have stated above, you might want to re-read my comment, I did say 'perhaps'. You are the one who said '2% silver solder causes this problem?'.

I personally don't have any skin in the game and other than my experience with silver plated wire, I don't claim to be an expert. And whether this is cuprous oxide or copper oxychloride, the OP's board is not in good shape. However, if you have a method to give to the OP to resolve his problem, please proceed, because I have merely hobby experience with boards and don't claim to be an expert. Again, if you are offended, this is not my problem. Or are you one of those people who always has to have the last word or wants to argue semantics and line by line disputings forever?

With all of that said, which is way too much in response, I have nothing else to say about this.

[helius]

Acrylic conformal coating is pretty good at excluding moisture, but it doesn't inhibit corrosion directly. There are conformal coatings that contain corrosion inhibitors (usually benzotriazole), like VpCI-286.

Benzotriazole forms a copper complex that prevents the oxidation of copper and formation of copper salts. It can halt the type of corrosion shown, although not reverse it.

[T3sl4co1l]

I don't care to call you on anything, merely asked your point.

It's regrettable that natural language is so bad at articulating a point precisely.

I don't mind writing it, but it also comes off as intimidating, or so overly frank as to be mean.  That's not my intent, but it's that much harder to convey such, while also staying precise.

Tim