Galvanic corrosion

[Pegasus79]

Hi there,

I am about to stack 3 or 4 feather boards with stacking headers and place the device near to sea-water, where it is expected to be exposed to water vapour and salt for really long time (10-20 years expected life time).
Such conditions benefit galvanic corrosion and of course i would like to try to reduce the risk of it as much as possible.
Here i found some advices how to pair metals to reduce this problem to occur

However i do not know which plating method/surface finish is applied to Adafruit PCBs. As a consequence i have a hard time selecting pin headers that have the most suitable plating material. In particular my main board is the ESP32S3 feather (ID 5323). My questions are:
- Does galvanic corrosion occur at all if the pins are soldered to the board? (i assume yes as the solder contains dissimilar metal as the plated hole and/or pin)
- If galvanic corrosion may occur on soldered pins, then shall i match the pin plating material to copper (metal that plates the drilled holes on the board) or rather to the metal used during surface finishing of the PCB?
- Are there any further considerations to prevent galvanic corrosion?

I guess this would be an interesting topic also for a blog post as people making only DIY most probably have no educated info on the topic.

[adafruit_support_bill]

Adafruit pcbs are plated with gold. But various types of corrosion are still possible in harsh environments. Methods to further inhibit corrosion include:

• Thorough cleaning of the boards
  Conformal coatings
  Protecting with an IP65 or similar rated enclosure

For more details, see:
https://vinatronic.com/blog/pcb-corrosi ... revent-it/

[Pegasus79]

Thanks for the tips and rapid answer Bill!

If i get it right, gold plating is the best option to prolong shelf-life of the bare PCB if the connection is purely mechanical (like gold plated edge connectors). However if the assembly contains any soldering phases, gold plating may get problematic due to amalgam building during soldering.

So best practice related to contact pins on gold plated PCBs used in demanding applications is either to use golden plated press-fit connector pins, OR in case the connector pins are fixed by soldering, then the gold plating should be removed from the PCB first by double soldering the pin holes. In the first round of soldering we just apply solder into the pin holes, then we remove the solder including the gold originating from the plating (that had the role to protect against surface oxidation during shelf-life of the bare PCB) and then during the second soldering round we should rather prefer using tin plated pins as by now we already removed the golden plating from the pin holes just some minutes ago.

I just do not know how important this can be in real practice? Or do you think following only the other points you mentioned would prevent quality degradation on the long run in those conditions? I mean if there might be a significant difference, then i am happy to double solder all the holes. However if this benefit is purely theoretical, i would not expose the PCBs to several heat-up cycles.

[adafruit_support_bill]

The issue with gold is embrittlement rather than corrosion. Stress risers can occur if the soldering process does not allow the gold to completely dissolve in the solder joint. For circuit boards regularly subjected to stress, these stress risers can lead to cracks and ultimately fracture the joint. And since this condition cannot be determined by visual inspection, the standard for critical applications is for the gold to be removed.

Of course, for such applications, the boards would also most likely be populated with components rated for extreme conditions rather than the standard commercial grade components we use on our boards. And the gold removal would need to be done prior to the pick&place of those components during manufacturing.

[Pegasus79]

Ok, thanks! Now i have a better understanding.

[adafruit_support_mike]

For the application you described, metal chloride formation and corrosion will be far more of a problem than gold intermetallics.

All of the metals used in electronics.. copper, tin, aluminum, gold, nickel-iron alloys like Kovar, resistance alloys like nichrome.. form chloride salts. Some of them are water soluble while others aren't. Copper is a prime example: Cupric chloride (CuCl2) is soluble at a level of about 7g per 10g of water. Cuprous chloride (CuCl), on the other hand, is only soluble at a level of about 0.05g per liter of water. CuCl decomposes to CuCl2 in the presence of oxygen, while CuCl2 forms two molecules of CuCl in the presence of metallic copper. The result is a lovely system that erodes any solid copper it can find and turns it into a white paste. It's one of the major industrial processes for etching copper.

Chloride ions congregate at any positive voltage, and can fully disassociate from sodium at 2.7V, basically becoming hydrochloric acid.

If you need a device to last 20 years, I'd suggest packing it in a grease like Cosmoline. Running electronics in an oil bath is fairly common in applications that require good temperature control, so as long as the grease is non-conductive the electronics won't care.

If you're near a coast it will probably be worth your time to talk to a good ship's chandler or carpenter. For any technical problem, it usually pays to find people who've spent their whole carreers dealing with similar problems.