has anyone measured performance of silver plating solutions?

[StuartA]

Back in the 1980's, I was involved in a detailed study of how best to pickle/passivate stainless steel to achieve optimum compatibility with the liquid rocket propellant dinitrogen tetroxide. After treatment by various methods, we looked at the surfaces with both an electron microscope (SEM) and a photoelectron spectrometer (XPS). Under the SEM, we could see that different pickling/passivation procedures resulted in different surface topographies. But when we looked at the chemical composition of the surface with the XPS, the composition hardly varied from one process to the next. Explanation - the final stage of each of the various processes was always the same; rinse off all the reactive chemicals with deionised water, while handing in air (~20% O2). Outcome, that final water/ air exposure has the final effect on the chemistry, we forget just how reactive water and air are and the test pieces all come out with about the same surface composition. We had a lot of fun doing surface chemistry, but time spent on the SEM and XPS is hard on project budgets!
S

[coppercone2]

Could you just confirm what materials these solutions are intended for?

It should work on all metals, but note that this will leave the surface activated. You need to passivate the work if no further processing is to be done.

DO you know if the mass of sodium carbonate here is anhydrous or typical hydrate (before you add the water, what are you weighing out)?

[helius]

 I believe it is given as the monohydrate. It shouldn't matter, though, there is only 18% between them. The carbonate acts as a buffer so it isn't critical.

[coppercone2]

Look forward to trying this soon. Can I try it with a sponge with similar results or do I need to setup a bath? I don't like chemical baths, mainly because I would need to make containers. All those chemicals are very cheap BTW, but it seems that you need to dispose of the phosphate correctly because its a environmental pollutant (another reason why I don't want to do a bath with large volumes, annoying to clean/evaporate/boil/etc).

(I don't want to spend 10$ to buy more polypropylene jugs LOL :0)

*interested in this in weld cleaning for stainless ATM. I know if you do a PCB you should probably make a tank.

Also, do you know what the benefit of the silicate is in the solution? I have it, but I noticed stuff I make up with silicate in it, i.e. etchant with silicate and NaOH, it crystallizes a puck of glass(?) in the container after a while. I assume it slows the reaction down and temperature stabilizes it?

DO you know how shelf life is effected with the silicate added?

And what do you recommend for the electrode?

[helius]

Trisodium phosphate is commonly used by painters to clean house siding, where it all washes into the watershed. It has minimal environmental impact and no special treatment is required. You aren't going to create a huge algal bloom with 100g of TSP...

You could use a carbon or stainless anode. Carbon would be better if you are going to electroplate the work because of the possibility of chromium contamination from SS.

https://orchid.ganoksin.com/t/electro-cleaner-for-pen-plating/34430
http://web.archive.org/web/19991010051525/http://www.metalworking.com/dropbox/E-CLEAN.TXT

Silicate is a flocculant, it attaches to debris particles and pulls them out of the bath. Probably it is less effective with brush methods. You're right that it may shorten shelf life.

Also check the forums at www.finishing.com, there's a great depth of experience there.

[coppercone2]

I wonder what a reasonable test would be to test the effect of the polish on a PCB.

If strip line was made on FR4 board, could a polished and unpolished piece be evaluated by high frequency S parameters without ridiculously sensitive and precise equipment? And, would FR4 variance be low enough so the data could be trusted?

Or would you need to etch it on ceramic PCB to eliminate dielectric variance? I think I read some where that things like PCB filters begin to be effected at 10GHz because of surface finish. I wonder what the least sensitive structure is that you could make to minimize etching inconsistency that would highlight the effect of surface roughness, in order to determine if the polishing does anything, and rule out other problems like minor mis-positioning of solder connectors/terminations, to verify if the process does anything to signal integrity from the standpoint of pure initial conidtion RF propagation (not corrosion resistance, longevity, or other factors that the polishing might improve from a 'mechanical durability' standpoint). So it would make sense to evaluate this on freshly prepared bare copper, or something coated with a electroless fresh tin strike (or something that is fast acting and puts a minimum layer on that will be consistant with exposure time only,(high process control).

[IconicPCB]

design a quarter wave resonator and look at the surface effect on Q.

Say three identical structures, one in as supplied copper laminate, one in polished finish and one in surface roughened finish , say a fine grit scourer pad.

[StuartA]

It should work on all metals, but note that this will leave the surface activated. You need to passivate the work if no further processing is to be done.

If those solutions leave the metal surface in an "activated" state which has to be passivated, what would you use on copper to do that? When you take the metal out of the cleaning mixture, presumably you have to wash the surface very well with de-ionised water to remove all traces of the reactive chemicals? What do you do after that?

[coppercone2]

I think its activated because its super clean. Thats why I suggested a tin strike right after you get it to the surface finish you want. I think you need to laquer or coat it with another metal.. I don't think you can deactivate it any other way. I assume a non-even dielectric coat of laquer would mess with it.

I think when you take PCB out of a ferric etchant tank they are also activated, the color right when its being pulled out of water is very unique.


The coolamp would need to be, polished after its done, if you follow the directions, because it requires scuffing during application. The other solutions use inferior silver.

So I think the choice would be a passivization layer that is ultra thin, and desposites insensitively to electric parameters (I don't think a filter will electroplate well).

Maybe jewelry dip silver plating if not tin?

Oh, and of course most PCB are just passivated with a silk screen, but since this project is working in the 10GHz regime, I am not sure how good of a solution that is.

For experiments, even if tin is RF bad, it might be a good way to analyze the surface finish, if its done quickly with a fresh electroless solution. I am not sure if there is any other way to get a uniform coating that does not effect the surface polish other then electroless metal plating. And you might need to re-clean that and let it reoxidize?

***
The other recommendation I would have is to get a air compressor and blow dry the part after rinsing with as high of a pressure as you can get to dry the surface out.

[coppercone2]

speaking of strip lines, I wonder what the surface finish on copper tape people use is. (fresh 3M). Maybe those tape filter prototypes have an advantage in this regard?

also , this might be how to monitor the current to determine process state


I think it would be the same for a copper PCB. He uses H3PO4, but I got the chemicals for the bath listed in the thread. Wonder what the difference is. It seems difficult to get the 'curve'.

Does anyone know if copper should have the same curve?

[StuartA]

I think activated means super clean, so, free of carbon (usually,'carbonaceous crap' covers pretty much everything in the World, which really is a farm-yard outside of a clean room) and free of copper oxide - but, as soon as you wash it with water in air to get the cleaning reactants off, I suspect that you'll form a very thin layer of copper oxide. We are talking about layers which down in the nm scale of thickness, but none the less, it is such layers that can make a big difference to a material, best example perhaps is stainless steel.

I've never worked professionally on copper, but there must be a mountain of info about preparing it for plating and actually doing the plating. One of the things we found was that trying to clean corrosion test specimens using organic solvents was poor at getting off the carbonaceous crap. For ferrous metals, our favourite cleaning technique for test specimens was to use a proprietary cleaner called Decon-90 in an ultrasonic bath, which cleaned very well, but did not etch. It became a standard procedure in the UK nuclear industry.

[coppercone2]

The good thing is those should not have an effect on microwaves. But I wonder at what frequency you start to get effects from that kind of oxidation in a way someone can measure at home.

This is what I have in mind when thinking about this problem
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.488.9435&rep=rep1&type=pdf

The only problem is, E8364B, at 60,000 USD on ebay. I was hoping those resonators that were suggested would make it more detectable (trying to cheapen the experiment so it will work with our zombies and swamp pull connectors).

I have a feeling detecting anything nm thick with AC (not DC leakage) is going to be very expensive from the standpoint of semiconductor/conductor physics. This has mainly to do with skin effect right now, I think what you are wondering about is more related to thin film optics (if we are talking about wave propagation not general quality)?

[IconicPCB]

We polish our copper laminate prior to electroplating.
We use a mixture of acid and IPA.
IPA provides the the insulating layer and reduces chances to excessively etch copper away.
There is also a characteristic current curve which resembles the first half of the curve shown in the video. Polishing done when the current drops ( due to alcohol in the mix ).
This is a part of the hole wall activation process which is based on copper hypophosphite which through pyrolysis deposits copper nano particles on the hole walls.

[helius]

Activated means that metal atoms are exposed on the surface of the work, not covered by a protective oxide layer. Metal must be activated in order for electroplating to work, because the atoms deposited from the bath need to form metallic bonds with the base metal. Oxide gets in the way and makes the plating weak and mottled.

Passivating is a finishing process that covers up the metal atoms with a protective layer, which can be oxide, nitride, phosphate, chromate, etc. Steel and aluminum should be passivated to prevent corrosion. With copper PCBs the usual procedure is to plate metal over the copper, such as with ENIG; solder over it, as with HASL; or cover it with stable flux-like corrosion inhibitor, OSP. Passivating the copper is not usually used because it would make soldering difficult.

[coppercone2]

you never gave the formulation of the phosphoric acid/isopropyl cleaner that you have good experience with

I see its also very cheap and useful for metal work.

What happens to electropolished vias?

[helius]

As long as the vias are mechanically sound they should be unaffected by electrocleaning, but their location may be too "shaded" to be effectively cleaned by the process. Ultrasonic will penetrate better.

It's also possible to use alternating polarity for electrocleaning as well as with electroplating. The anodic phase will repel any loosely adhering metal atoms that got pulled onto the work during the cathodic phase, leaving a smoother finish. It's done for only a few seconds to avoid pitting the surface.

[coppercone2]

hmm, is there any application of AC balance here (dc offset)? and what do the frequencies look like this process?

[IconicPCB]

Activation deposits copper nano particles over all surfaces.

Nano particles are embedded into the de-smeared epoxy surfaces of the hole walls.

Nano particles are therefore desirable inside the holes as they for seed layer for electroplating.

On the flat outside surfaces they would form a rugh finish plating if not removed ( polished off).

So electropolishing prior to plating ensures a level smooth plating finish.

[coppercone2]

Is there a hack to use voltage here? Like the nickel plate I have specifies a Current/Area, but it also says its just acceptable to use 2V and let it operate as powered by a voltage source.

I.e. the range it gives is 10-20A/Sq-foot, or 2V (well a high current 2V but still). Makes it a little easier.

Feels like its going to be difficult to figure out the current density you need unless you calculate the cad files area. Not a big deal unless you have old masks and no gerbers, or just image files saved.

[StuartA]

The amount of info on nickel plating on the internet is all but infinite, but you do see some variations in the conditions recommended, often in regard to the temperature and current density to be used. Of course, there are a number of different "recipes"for nickel plating solutions.

Possibly better still, if you could source some, are the electroless nickel plating solutions. As the name implies, you don't need to apply a current, they just deposit nickel out of solution. I've used one proprietary product in the past and it was truly outstanding, but the company which supplied that is no longer trading. Unfortunately, this tends to be the kind of stuff where the minimum quantity they will deal with is a drum containing a few gallons.

[coppercone2]

lol, I tried this on a small scale, with a HP 3A linear supply, a sponge, the solution without silicate, stainless steel electrode. I put a slab of stainless steel inside of a glass tray and put a sponge on top of it and then poured a bit of solution on it.

I attached the other wire it to a dirty ass copper scrap (small ~0.7mm thick sheet cutting that looks like a small slice of pizza) and it turned black in a pattern resembling asphalt after 30 seconds after being pressed into the sponge.

Just a warning, you appear to need a half decent setup for this procedure. Or the temperature actually matters. Not sure if you can heat it up on a hot plate to 70C and dunk it once in a while if you are trying to work without a bath. And you probably need a sponge thats uniform, like a nice soft one, not the 'lava rock' like thing I have. I heard sizzling.

I don't want to work on this until I setup a regulated battery powered power supply (dislike using mains equipment near pools of water, it makes me nervous. I try to minimize the number of things that are plugged in near ultra conductive salt solution)

And it might also need to be cleaned from oxides first, the stuff I tried was the grungiest piece of copper you will see. Perhaps it needs to be buffed a little first to bring it to a copper like state. Good quality copper is a rarity because I use it up quickly.

oh i know, i will try a penny

[coppercone2]

On a penny in submersion it works well, it came out with new shine. The sponge needs a greatly improved apparatus.

BTW its cathodic, the cleaning process shown on youtube is anodic with phosphoric acid in most cases

If you do this on a PCB it would need to be carefully jumpered so you can machine the jumpers out of the circuit, not trivial at high frequencies that require precise PCB lengths. Maybe a small diamond burr would work. I noticed things cut with diamond have much nicer surface finishes in the same grit (i.e. diamond lapping plate leaves it looking like its ready for a metalurgical inspection if you do it on something, say a driver bit side).

Does anyone have an idea about how to link up a complex PCB for electropolishing on the signal layer? Unless you do the photoresist yourself, its going to be difficult. Especially with SMD stuff. A ground plane is much easier. Doing a bunch of cuts is very labor intensive and adds alot of work to layout. 

Granted, unless you are doing a full on portable miniature spectrum analyzer, individual cans are not that bad. Things like DC routing can be ignored. And it would only be really beneficial in the LNA i think. Active antenna amp? I am suspicious that sticking the PCB on a buffing wheel with fine (green) jewelers polish (chromimum oxide) might have better results too.

[coppercone2]

what the hell is this solution doing. I put a wave guide piece in there (brass) , in a tub, partially submerged, the part facing the stainless electrode turned carbon black. The part facing away from it got polished impressively.

[coppercone2]

you can see the submersion level, one is the copper pipe the other is a waveguide brass piece.

The black line is the water level line. Some of brass I cleaned up on the wire wheel, it does not get nicely polished like that. I wonder if maybe its malfunctioning because the waveguide couplers are made of a different material then the waveguide body.

Why is it turning my waveguide black but working on copper? not suitable for all metals? something with tank /bath parameters?

[helius]

zinc and chromium ions reacting to form a chromate conversion material?