[wuhtzu]

Hey everyone

I have a few questions / thoughts about galvanic corrosion of Ni (nickel), Cu (copper) and stainless steel which I will like to discuss - the whole thing in relation to the chemical / physical aspects of water cooling systems (for electronics):

The cooling system
This is a schematic diagram showing the components in the cooling system:
http://wuhtzu.dk/random/water_cooling_diagram.jpg

Components:
  - Hose fittings made of Ni (nickel)
  - Heatsink made of Cu (copper)
  - Pump made of 316 stainless steel
  - Hoses, containers ect. made of plastic


Basic problem
The system is comprised of metals with different values of electronegativity in contact with an electrolyte in form of water with possible additives.

  - Will galvanic corrosion occur?
  - Which metal will corrode?
  - What can be done to stop the corrosion?

Some data:
Electronegativity: http://en.wikipedia.org/wiki/Electronegativity
Mn(1.55) < ZN(1.65) < Cr(1.66) < Fe(1.83) < Cu(1.90) < Ni(1.91) < Mo(2.16)

316 Stainless steel: http://www.azom.com/details.asp?ArticleID=863
Fe, <0.03% C, 16-18.5% Cr, 10-14% Ni, 2-3% Mo, <2% Mn, <1% Si, <0.045% P, <0.03% S



#1: Copper in aqueous copper sulfate
CuSO₄ (Copper(II)sulfate) is often used as pesticide in water cooling systems. Will this have any impact on the copper heat sinks?

What will happen to solid copper in aqueous copper sulfate?

  Cu_(s) + Cu²⁺ + SO₄^2- -> Cu_(s) + Cu²⁺ + SO₄^2-

Nothing should happen, right? The only other thing I can think of happening would be electron transfer bewteen Cu_(s) and Cu²⁺ causing the "original solid copper" to degrade and "new solid copper" to form elsewhere in the solution. But it doesn't sound plausible to me. What do you think?


#2: Copper(II)ions can corrode chromium and iron
Cu²⁺_(aq) has the potential to corrode Cr_(s) and Fe_(s) which the 316 stainless steel is made of. Will that happen?

Cu²⁺ + Fe_(s) -> Fe²⁺_(aq) + Cu_(s)

They are not in metallic contact with each other (the copper and stainless steel). They are only connected via the aqueous CuSO₄.



#3: Nickel ions can corrode copper
Ni²⁺_(aq) has the potential to corrode Cu_(s):

Ni⁺²_(aq) + Cu_(s) -> Cu²⁺_(aq) + Ni_(s)

Since there is no nickel ions to begin with only solid nickel in form of nickel plated fittings the above reaction shouldn't take place, right?


#4: A sacrificial anode
Will it do any good to add a sacrificial anode in form of a solid zinc Zn_(s)?

In theory the Zn_(s) should corrode completely before any other metals start corroding.




What do you think about the above thoughts / problems. Am I somewhat right about my assumptions or am I way off?


Any thoughts / answers to my questions will be greatly appreciated
Wuhtzu

[Arkcon]

The standard additives to automotive coolant added to your cooling water (distilled or deionized, naturally) should protect all parts from corrosion.  That said ...

A galvanic cell can only form where two dissimilar metals are in contact, usually, the biggest problem is iron and copper.  Nickel and copper, the only dissimilar metals in contact in your diagram, might react, but may be close enough in electronegativity for the reaction to be so gradual as to not be a problem.

Added copper ions will gradually plate out as copper metal on iron.  Steel may be immune.  But it isn't as good at protecting the water from microbes as the above mentioned additives are.  Copper II sulfate does not attack copper metal.

Sacrificial anodes are for ships that have regular maintenance periods, during which they're replaced.  Meantime, they will obstruct the liquid flow.  And are you really going to open the pump and weld a zinc bar in place, monitor its dissolution, and replace it when needed?

[wuhtzu]

Thank you for your reply Arkcon.

If two solid metals have to "corrode" each other will they have to be in metallic contact for the electrons to transfer?

For example will Cu_(s) need to be in metallic contact with Ni_(s) for corrosion to occur?

Because if that's the case then there should be no problem since non of the metals are in metallic contact with one another. They are all separated by plastic.

[Arkcon]

That's correct.  If there's some sort of plastic between the Ni and the copper block, you've got nothing to worry about.  Look up the story of The Statue of Liberty -- an iron framework with copper skin.  Two tricks were used: corrosion resistant iron, and asbestos/tar pads to separate the two.  In the end, only diligent maintenance is the real answer to corrosion.  That is -- inspect frequently, and replace.  Regular fluid changes, if the fluid contains additives to inhibit corrosion, are the answer.

[wuhtzu]

Thanks again Arkcon.

...
Added copper ions will gradually plate out as copper metal on iron.  Steel may be immune.  But it isn't as good at protecting the water from microbes as the above mentioned additives are.  Copper II sulfate does not attack copper metal.
...

The above sounds like the iron will not be corroded, just plated with copper. You are referring to this reaction, right?

Fe_(s) + Cu²⁺_(aq)  ->  Fe²⁺ + Cu_(s)

If I add Zn_(s) to the above reaction will only zinc be oxidized or will both zinc and iron oxidize?

(1) Zn_(s) + Fe_(s) + Cu²⁺_(aq)  ->  Zn²⁺_(aq) + Fe_(s) + Cu_(s)

or

(2) Zn_(s) + Fe_(s) + 2Cu²⁺_(aq)  ->  Zn²⁺_(aq) + Fe²⁺_(aq) + 2Cu_(s)

This is what I referred to as "sacrificial anode" in my original post.

In laymen's terms, it's a piece of readily corrodible metal attached (by either an electrically conductive solid or liquid) to the metal you wish to protect. This piece of metal corrodes first, and generally must dissolve nearly completely before the protected metal will corrode (hence the term "sacrificial").

According to the above quote, reaction (1) should occur until all zinc has been corroded. If that is the case, that zinc will corrode before the iron, over what distances will this work?

[Arkcon]

For a zinc sacrificial anode to work, it must be welded to the iron it is to protect.  I don't know what surface area it will extend to, but ships' hulls that have sacrificial anodes attached need several.  They are replaced as ions in solution consume (by corrosion) the zinc.  This phenomena is not going to help if there is Cu2+ ions in solution, they will rapidly drop out as Cu0, and the zinc will rapidly disappear.  Better to forget it, and use the store-bought anti-corrosion additives.

[wuhtzu]

I have realized this is not going to work... just to let you know I'm not stubborn / stupid I tested chromium, iron, stainless steel and brass in aqueous copper(II)sulfate and they corrode at an alarming rate. So there is no way you want to use it in a water cooling system with other metals than copper.

Now (and maybe from the beginning) it is just to understand some more chemistry.

But thanks for your replies Arkcon

Wuhtzu

[Borek]

Can you describe your experimental setup? Copper sulfate concentration? Observations that lead you into thinking metals corrode?

[wuhtzu]

Sure Borek.

I'm not in a lab these days because of Christmas, so I only have very basic equipment at my disposal, but it should be good enough to give estimates about e.g. concentrations.


The solution I am 'testing' with is made by dissolving CuSO₄ · 5H₂O (Copper(II)sulfate pentahydrate) in normal tap water. It has an approximate concentration of:

Molarity: 0.2M
Weight percent: 3%

Almost immediately the metal objects start to corrode and after a few hours all Cu²⁺_(aQ) is used up and the metal objects are covered in a thick layer of solid copper. Most of it dropping off when removed from the solution.

Here is some pictures:

The Cu²⁺ solution:
http://wuhtzu.dk/random/corrosion_copper_sulfate_bath.jpg

Brass:
http://wuhtzu.dk/random/corrosion_screw_cover.jpg

Zink plated:
http://wuhtzu.dk/random/corrosion_screw.jpg

Unknown:
http://wuhtzu.dk/random/corrosion_carabiner.jpg

Chromium plated:
http://wuhtzu.dk/random/corrosion_hose_fitting.jpg


The reaction occurring must be:

Me_(s) + Cu²⁺_(aq)  ->  Me²⁺_(aq) + Cu_(s)

where Me represents a metal.

[Rabn]

I don't know about all this...I'll list my thoughts:

If there are ions in the water it is the same as having the dissimilar metals in contact because the ions will carry charges between the metals.

What you need to look at is the reduction potential of each metal, if they are close enough you shouldn't have to worry about corrosion with the proper additive.

In that kind of setup you should try to match the metals wherever it is possible. If there are dissimilar metals they will eventually corrode, it is just a matter of how much time it will take. 

In your case the amount of copper ions in the water will determine how corroded the Nickel gets.  I would try to replace all of the fittings with plastic fittings, they work just as good and corrosion is avoided.

[Borek]

If there are ions in the water it is the same as having the dissimilar metals in contact because the ions will carry charges between the metals.

It doesn't work this way. Besides, there are two separate processes here. One is reduction of copeer ions on the metal surface, second is redox reaction between two metals being in contact. Reaction takes place on the metal surface that is in contact with water, but if there is no contact between metals reaction doesn't take place.

[wuhtzu]

Since no two metals are in metallic contact with one another the only reaction taking place must be the  reduction of copper ions. But as you can see it's pretty severe.

[Borek]

Since no two metals are in metallic contact with one another the only reaction taking place must be the  reduction of copper ions. But as you can see it's pretty severe.

I must tell I am pretty surprised by the extent of the reaction.

[enahs]

I am a little confused about the setup? Are you getting these problems in an actual cooling rig, or are you just experimenting with the various metals in a small container sitting in tap water?


I mean, that sound like to me, if it is the case, just a very complicated battery/galvanic cell. You have more then enough to form various redox couples, a pseudo salt-bridge can be created, and electrolytic solution, as well as electron transfer.

I suspect, if the case is you are just sitting the metals in still tap water in the same container, just putting a magnetic stirrer in and keeping it moving will reduce the pseudo salt-bridge and redox complex and slow the reaction?

[Rabn]

If there are ions in the water it is the same as having the dissimilar metals in contact because the ions will carry charges between the metals.

It doesn't work this way. Besides, there are two separate processes here. One is reduction of copeer ions on the metal surface, second is redox reaction between two metals being in contact. Reaction takes place on the metal surface that is in contact with water, but if there is no contact between metals reaction doesn't take place.

That isn't how it works? It's a galvanic cell, the electrodes, being the metals, don't have to be in contact. I'm speaking from experience here as well theory. Electron transfer will occur between the metals just like electron transfer happens in the atmosphere...moving molecules. You are right, there are two different redox couples happening. The metal/metal at the junction between the fittings and either the cooling block or the pump(assuming that the fitting is in contact with the steel; as well as the galvanic redox of any copper ions in the cooling solution. From my experience in semi-conductor manufacturing, where we used various methods of temperature regulation, with the set-up you posted as long as you use a 50/50 mix of distilled water/ethylene glycol you won't have to worry about corrosion. I'm assuming this is for a cpu cooling rig; by the time you are ready to upgrade to a new socket and require a new cooling block the only corrosion you'll have will probably be between the fitting and the copper block, which I would say is desirable because it would seal the fitting against any water leakage through the threads.