MEK to Remove Solder Mask from PCB

[AJ3G]

All:

I have some 1.5” Square Boards PCB’s which I want to strip all the solder mask off.  I have heard MEK can do this, but how well is unknown. Has anyone tried using this solvent for stripping solder makes. If the results were horrid, I would like to know. 

Rich

[graybeard]

MEK attacks almost anything organic.  I would worry about it attacking the binder in the PCB. 
I have used it to clean deposits in antique motorcycle engines, but never on a PCB.

[magic]

It may or may not. FR4 is some sort of epoxy resin and usually quite chemically resistant.
I came across papers about dissolving waste PCBs the other day. They used aggressive organic solvents like DMF and NMP at elevated temperatures to do the job.

[mikeselectricstuff]

I think anything that will dissolve resist will also dissolve the resin in the substrate, though a lot more slowly due to the increased thickness, so may be viable if it's a 1-off job for reverse-engineering etc.
A PCB left for a couple of weeks in methylene chloride/dichloromethane will end up as a pile of copper and glassfibre matting. 

[magic]

Did you try this exact experiment or are you extrapolating from something else?

I have never done that, but I have tried soaking ICs in DCM for a week to dissolve them and it did nothing. Those epoxies used in electronics are very resistant. Unlike some others, which fall even to acetic acid.

[mikeselectricstuff]

Did you try this exact experiment or are you extrapolating from something else?

I have never done that, but I have tried soaking ICs in DCM for a week to dissolve them and it did nothing. Those epoxies used in electronics are very resistant. Unlike some others, which fall even to acetic acid.

IC packaging resin is completely different from the stuff used for PCBs AFAIL only hot nitric acid wil dissolve IC packages

[fcb]

MEK is wicked stuff - our local screenprinter who died in November from lung-cancer swore by it.  Apparently it's not carcinogenic.

[T3sl4co1l]

MEK is a bit more aggressive than acetone.  Not nearly aggressive enough to affect epoxy inks and resins.  Neither is THF; many conformal coatings happen to be susceptible to it though.

N-methyl-pyrollidone and DMSO may be less toxic (but more expensive..?) alternatives to DCM.  Both absorb readily into the skin...

In any case, the substrate is epoxy too, so you can't let it soak, you'll just turn the whole board to mush.  I would guess, wiping on some solvent, letting it soak in slightly, then abrading away that layer, and repeating until the mask layer is clear, is the way to go.  Fiberglass brush is usually recommended.

Also, frikken lazers.  Won't harm the copper, of course again you need to avoid damaging the substrate too deeply in the process.

Tim

[magic]

If soldermask is made of epoxy like FR4 than things may be tricky indeed.

Although the truth is that there is actually no such thing as "the epoxy resin". Epoxy is not a particular chemical like polyethylene or PTFE but really a process for polymerizing various kinds of molecules into heavily cross-linked networks, named after the epoxide group which is involved in curing of the resin and then is gone from the final product.

So the constituents determine what the "epoxy" really is and its chemical properties. If you are lucky, you may find some solvent which will remove one epoxy resin from another. Case in point, DCM (supposedly) will eat PCBs but will (certainly) leave transistors and ICs which are "epoxy" too.

[T3sl4co1l]

Indeed, it's a diverse system.

FWIW, semis have the distinction of a large loading of silica fume or the like.  Hence the high arc resistance rating, high ash content, stiffness, etc.  Some chemical ideas here: https://www.pcimag.com/articles/104416-epoxy-resins-for-semiconductor-packaging-assemblies

Tim

[wraper]

If soldermask is made of epoxy like FR4 than things may be tricky indeed.

Although the truth is that there is actually no such thing as "the epoxy resin". Epoxy is not a particular chemical like polyethylene or PTFE but really a process for polymerizing various kinds of molecules into heavily cross-linked networks, named after the epoxide group which is involved in curing of the resin and then is gone from the final product.

So the constituents determine what the "epoxy" really is and its chemical properties. If you are lucky, you may find some solvent which will remove one epoxy resin from another. Case in point, DCM (supposedly) will eat PCBs but will (certainly) leave transistors and ICs which are "epoxy" too.

Normally solder mask is negative photoresist. Either liquid or dry film. After exposure to UV and baking at high temperature it becomes very durable.

[magic]

I found this PCI article too a few weeks ago but now it's showing me a paywall

Anyway, the most important takeaway I remember is that ICs are typically packaged in novolak based epoxies. Novolak is apparently a phenol-based polymer akin to bakelite. A quick search for chemical attacks on the latter brought me to the Science Madness forum where the general response was "no way" and "you have to be nuts", but somebody linked lecture slides by one dr. Hazizan which suggest that phenol resins may be vulnerable to 1-naphthol, formic acid and strong oxidizing acids. The latter of course is no-shit-Sherlock for anyone interested in decapsulation

But I'm digressing. No idea what soldermasks are made of, but perhaps you could find articles similar to the above about ICs.

I'm not sure if you will want to play with all those organic solvents once you read the safety datasheets, but formic or acetic acid seems safe enough to try. Even the mineral acids don't look too bad in comparison, at least not carcinogenic. They do nothing at room temperature BTW, so if you are lucky perhaps one polymer will give up before the other on heating. Maybe.

[T3sl4co1l]

I mean... acetic acid has a complex for copper, it might dissolve that faster.   And it looks formic does as well.

Tim

[Yansi]

Neither does MEK dissolve a common soldermask types, nor does it dissolve the PCB binder. Certainly not within a short application time. Bathing the pcb in MEK for extended periods of time may make a difference.

[Yansi]

MEK attacks almost anything organic.  I would worry about it attacking the binder in the PCB. 
I have used it to clean deposits in antique motorcycle engines, but never on a PCB.

Not true. MEK is a bastard, but certainly not "attacks almost anything organic".

Sure, your fingers and ABS plastic it eats for breakfast, but a common polypropylene? Maybe even polyethylene? PTFE? Likely doesn't touch it with a long stick.

[magic]

I mean... acetic acid has a complex for copper, it might dissolve that faster.   And it looks formic does as well.

Why is nobody etching PCBs in vinegar then?

It's not actually that aggressive against copper. I tried it once and got slightly bluish solution after a week of soaking and nothing visible on copper. Don't remember if it was 10% or 100% acid. Actually, 100% concentration may be weaker against metals if it behaves anything like nitric/sulfuric.

It will get worse with added salt, however. This seemingly innocent mixture can even etch aluminium in a matter of days.

[T3sl4co1l]

Vinegar is weak (5%).  I didn't say it would be fast, but we're talking epoxy here, so, there's probably some time for it to happen.

OTOH, hot glacial acetic acid, isn't very pleasant to work with.  Keep that away from my mucous membranes, please!  Given some oxygen (or H2O2), and indeed chloride and other ligands will help further, and you're likely to see some effect.

Tim

[SeanB]

Hot tricholrethane is very good at stripping the soldermask, and after around a minute more you start to lose the board as well. Was the only way to remove some of the more robust conformal coatings, though it would wreak havoc on anything with a rubber seal, and would remove all markings from components, making all the resistors and glass encapsulated components a uniform colour, though if you left a resistor in too long it came out with shiny metal end caps and a very easy to see carbon or metal film.

Did nothing to epoxy that IC's are made from, though we did lose a fair number of ceramic package lids in it, all of them the cemented type. Soldered and glass frit survived no problem.

[AJ3G]

I thought this may not be as straightforward as I thought. I was terrible at Chemistry in University.  I looked at some videos, ad I saw someone using a teaspoon of lye, and boiling water to accomplish the job. 

What I have done is purchased 4 of the boards, and will try a few approaches before doing the one I want to keep. The boards are $1.50 a piece, and no, I am not mining for precious metals, but intend on using the board but need the mask removed to implement/modify the board for my application. 

Typically I would just have some boards made, but I cannot purchase the boards from any vendor after designing them for the price this vendor is dumping a board that will work.  Kind of a odd request I know..

[magic]

OP gave up so let's hijack the thread

I got some 85% formic acid for other purposes and tested what happens to a piece of PCB immersed in it.

Three days later:
FR4 soaked it up and increased in thickness from 1.2mm to 1.55mm
Soldermask is easily scratched off from copper but still adheres to the laminate.

A small trace in the corner of the sample separated from the substrate and the soldermask. Copper wasn't appreciably dissolved, but one side carries some dark residue. No obvious loss of mechanical strength in either of the separated pieces.

Effective but not without side effects

[rdl]

These days almost all industrial coatings are "cross-linked" or polymerized in some way and are very resistant to solvents once they have reacted. Used to be it was referred to as thermoset vs. thermoplastic, but modern curing mechanisms often require little or no heat (UV for instance). The only significant industrial (oem) use of "thermoplastic" coatings I can think of off the top of my head are nitrocellulose lacquers for musical instruments, primarily guitars.

[AJ3G]

I did not give up, but Lye has worked for the boards I am working with. Feel free to take the thread, I will be glad to see where this goes.

[duak]

Paint stripper containing Methylene Chloride took off the solder masks I tried it on.  Using a non-polymer brush that won't dissolve in the paint stripper, a uniform coat is applied that foams up and holds the active stuff and crud in place.  I've found the process to be quite controllable.

Most "safe" strippers are lye based (Sodium Hydroxide - NaOH) eg. 3M's Safest Stripper probably won't touch the solder mask but will do a number on any metal.

Trichloroethane also works well but is hard to get in small quanities.  It also cleans off most inks.   Dry cleaners often use Perchloroethane but I don't know if it works on solder masks.

[magic]

Another piece of the same board went into DCM. Five days later thickness increased to 1.8mm, rigidity decreased and I could relatively easily separate it into layers.

However, solder mask is still intact and adheres well to copper. Silkscreen too
Maybe the DCM paint stripper contained something more or we have different solder masks.