• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Very poor article (Watercooling and electrochemistry)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I'm no scientist.. and I really don't care if I'm wrong.

But putting a chunk of anything in your reservoir or whatever, sort of implies the notion you are going to end up with rusty water.


My water wetter has never failed me yet, so I'm happy.
 
K2
you are asking a question about which most know little, self included

I was/am under the impression that only pure magnesium will be effective as a sacrificial anode,
so - if such is true and that is an error - I would be inclined to question the rest of the article

just what is 'poor' about the article ?

be cool
 
I am glad you asked, i have written a couple of things that i disagree with. The problem with corrosion is that it is a not an exact science, it is very hard to predict if corrosion occurs or not. Thermodynamics will give you an indication of what is possible and not, but that doesnt mean that a thermodynamically feasible reaction will actually occur in the real world. Here goes:

1. Aluminium does not dissolve into water !. The standard reduction potential is correct, but that only tells us about thermodynamics, not reaction kinetics. Even though aluminium is thermodynamically very reactive, it is in fact very stable in aqueous solutions, with pH values between 5 and 8. It is a known fact that aluminium’s primary corrosion product Al2O3, has the exact same crystal structure as pure aluminium. This means that newly formed Al2O3 will cover the pure aluminium and shield it form further corrosion. This protective layer is extremely thin, but it is sufficient to make aluminium fairly corrosion resistant. Chloride (Cl-) will destroy the protective layer, therefore it not recommended to use tap water in water-cooling. In some parts of the world they even add chlorine to the tap water to kill bacteria.

2. Electrons are not soluble in water !. Even if aluminium did not have its protective coating, the reaction would stop fast as the remaining aluminium would become negatively charged. The same goes for copper . The electrons need to “go” somewhere. They could be “used” in a reduction of an ion in the solution, but if you use distilled water in your setup there are no positive ions to accept these extra electrons, so again no corrosion (pure water does contain H3O+, but in very small amounts ~10-7M).

3. Zinc does not help in any way !. Zincs standard reduction potential is -0,76V, meaning that, points 1&2 aside, it will only worsen the corrosion. Zink is used, and works, as a sacrificial anode in seawater, where conditions are very different, but in clean water it will not help in any way.

4. Galvanic corrosion
First of all there are a couple of conditions that need to be fulfilled for galvanic corrosion to occur.

1. Electrical contact between to dissimilar metals.
2. An electrolyte.

I will not go into the reactions, as they are specific to each system according to the electrolyte. In a normal water cooling setup, a copper water block and an aluminium heater core are separated by hoses. Hoses (PVC or silicone) are very poor electric conductors, therefore there is no electrical contact between the to metals, and hence no corrosion. If you use distilled or demineralised water, the electrolyte concentration is so low that you need not worry about galvanic corrosion.
That does not mean that galvanic corrosion does not occur at all. If a small chip of copper is loosened from the copper block, gets stuck in an aluminium radiator and you are using tap water, you will see galvanic corrosion. The corrosion rate is very hard to predict, but it will most likely be very low. Galvanic corrosion in water-cooling is a myth, nothing else.
 
I think you're ignoring, or glossing over, several items

the observed corrosion of Al is not surface oxidation, it is in the form of a 'cell' (many in fact) pitting at a rapid rate
(now you can tell me that there are flecks of copper there but you would have to cite some sources to be believable)

and sorry, but water - distilled or DI (which is DI for all of 10 min in the presence of Al and Cu) - is a VERY serviceable electrolyte

and how can you be assured that the wb is NOT grounded ?
you can bet the rad is

be cool
 
I think this is why Aluminium-alloy 6061 I believe includes Magnesium, Zinc, Copper with the bulk of it being Aluminium. Aluminium-alloy 6061 has a good corrosion resistance because of this but I wouldn't say that its completely corrosion resistant although it's better than pure aluminium. The Magnesium, Zinc, Copper in it "I think" acts like the sacraficial anode.

BillA is right about the Magnesium as a sacraficial anode.
Also it's not the "copper flakes" as you would say that cause the corrosion. It's the copper ions that react to Aluminium that do the corrosion. I briefly read about the subject so I can't remember right off hand the science behind it all.

I made my block from 6061 and from 2 weeks sitting in regular tap water I havn't seen any corrosion. The brass barbs attached to the lexen top though seem to have suffered some but it doesn't look like corrosion but maybe oxydation. The insides of the barbs are a little darker looking than when they were shiny brand new. The inside of my crappy core still looks good.
 
Pure water is not conductive.
DI water by definition is not an electrolyte.
Proving this would require a $25 test bench, where $20 would be spent on DI water:)
 
The two metals have to be touching is how I understand it. Just as the Al block I seen ( I think it was a senfu) that corroded greatly which can be attributed to the barbs screwed into the Al block.

Here's two links from Nasa and here .
 
Last edited:
Welcome K2.

Sounds like you have some expertise in these matters.

I believe it's quite likely the article has flaws. As it was first written, Zinc was recommended as a sacrificial anode for aluminum. After several of us brought Zinc's inadequacy to the author's attention it was corrected. (I believe Caffinehog initially picked up on this and pointed it out.)

If you have suggestions for making the article more accurate, I suggest you contact the author. (There should be an email link at the bottom of the article.)

Galvanic corrosion has been seen in many CPU watercooling circumstances. Particularly in these cases:

Mixed copper and aluminum waterblock.

Aluminum block with brass barbs.

Copper block with aluminum barbs. (Don't ask me where someone got aluminum barbs and the bright idea to use them.)

A thread on this is here.

I don't know of any reliable reports of aluminum radiators corroding. However, as BillA points out, frequently the radiator is connected to ground. Also, some people have connected their waterblock to ground and reported higher overclocks. (I'm somewhat skeptical of these higher overclocks, but as an Electrical Engineer, I have seen a lot of bizarre and unexpected behavior, so I'm keeping an open mind about the idea.)

Anyway, it is quite possible that someone may have an electrical connection between their copper waterblock and aluminum radiator. And, it is quite common for people to use coolants that promote galvanic corrosion. Some have even added bleach to their water as a biocide.

I hope you will participate in educating people here on these issues. (I try, but I've had about 20 years to forget all the chemistry I learned in school. I often can't speak with confidence on these issues.)
 
all of these are good points, but i will still continue to put in an zinc anoide in to my res, and i have never had a problem with them or my system corroding around me, i got the idea from my boat, if it will protect my prop, why not my water block, it is cheap and i dont know if it works or not, but i dont think it will adversly affect anything, to have this zinc anoide suspended in the middle of my res.
 
Al rads certainly WILL corrode, just like the Al wbs
I have no experience with an all Al system

I do not understand, but I am no chemist by a long shot, how one can say that water is not an electrolyte (a solution that conducts electricity)

and while the resistance of DI water is measured in Megohms, this value plummets immediatly when in contact with Cu or Al
(I used to be a QC/QA supervisor in a wafer fab plant, trust me, DI water is a b*tch to maintain)

galvanic corrosion in CPU WCing systems is no myth

be cool
 
Maybe someone should objectively do detailed experiments regarding this. Personally I like the idea of applying reverse voltages on the electrodes.

I do know for a fact though that what k2 said about aluminum only oxidizing to a certain point is true, which leaves me confused why people get corroded al in systems. Maybe anodizing al would help, aluminum oxide is an insulator. Or is there a metal that will commonly plate onto al and cu.

SysCrusher, the additives to 6061 are mainly such as to improve its properties for mechanical purposes. I.e. tensile strength, wear resistance, machinability, weldability, friction coefficient, etc. the additives won’t be in sufficient amount to be much good as a sac. Anode
 
Water is not an electrolyte because the concentration of the ions in the pure water is small to negligable. To make it a sufficient electrolyte one would need to apply a voltage higher than that of breaking it down into 02 and H2.
This is an industrial proccess for super-clean O2 production.

Upon contact with something that is able to produce ions such as metals and more importantly metal oxides the DI water would certanly become contaminated (and stop being DI), but otherwice DI water it pretty good insulator.

Now in our typical systems the innards of a typical setup would hardly cvalify for a sterile ranking. Metals, metal oxides, remnants of the acid flux used in soldering etc. would create a pretty good electrolyte hence making the block grounded through the rad.

The pit effect observed is imho due to the use of aluminum alloys rather than pure metals. Protective layers of oxide are formed on the surface except in the areas where other metal's clusters present. At this point a galvanic corrosion takes place and aluminum acting as anode dissolves.

As per sacrificial anodes the series of metals in the NASA article linked above would serve as a guide. Theoretically any metal to the bottom of the serries (more active) below aluminum would act as a sacrificial anode. Thus both zink and magnesium would be ok.
 
BillA:
My main source is a Danish book “metallurgi for ingeniører” (metallurgy for engineers), that has a large section on corrosion. I don’t expect you to understand Danish, so that doesn’t help you much. I am a chemical engineer, and have used much of my time on metallurgy. That doesn’t make me a corrosion expert, but I know a little.

You are absolutely right that the observed corrosion of aluminium is often localized pitting and not uniform, this type of corrosion is referred to as pitting and not galvanic corrosion. If in fact you did have galvanic corrosion, it would be uniform, spread out over the entire aluminium surface, that is the very nature of galvanic corrosion. Pitting is a different type of corrosion, not involving 2 metals but only 1. Aluminium is especially prone to pitting, because of the protective oxide layer. Here is a good reference: http://www.eaa.net/transportation/corrosion_pitting.asp

Pitting (aluminium) is a far greater concern for water-cooling enthusiasts than galvanic corrosion. You must avoid chlorine ions (Cl-) if you have aluminium in your setup, that pretty much excludes tap water, especially if your local water supplier adds chlorine to the water to kill bacteria. I think http://www.dansdata.com/burning.htm is a textbook example of pitting.

Galvanic corrosion can easily be avoided by making sure your rad and/or WB is not grounded. You could mount your rad with rubber washers, no galvanic corrosion. Distilled or DI water are both poor electrolytes, as walkerIV explained, that will also impede galvanic corrosion. Personally I have been using a copper wb and aluminium radiator for 6 months without problems, and will continue to do so without loosing sleep. I use DI water and 15% antifreeze to keep algae and bacteria out.

SysCrusher
True, Aluminium-alloy 6061 is a little more corrosion resistant than pure aluminium, but for different reasons than the one you mentioned. The alloying elements help stabilize the protective oxide layer, because they influence the crystal structure.

You mention that copper ions are responsible for the corrosion, but where do they come from ?. If you put a copper block in distilled water, there are no or only trace amounts of copper ions in the water. Copper does not spontaneously decompose into ions and electrons, if you have read otherwise, I would be very interested in reading it, as it basically contradicts what I have learned.
 
Last edited:
interesting stuff here

ok, so we do have an electrolyte - but cell corrosion rather than galvanic
sounds eminently reasonable to me

and the cell corrosion is related to impurities (i.e. non-Al) in the metal
ok
-> so it then follows that were PURE (5 9s ?) Al used and the wb not electrically connected to any other metal,
no corrosion would occur ?

I wonder if such is worth the cost or effort
pure Al is not so easy to obtain and I understand its difficult to machine
all this for lower performance ??

off topic
K2, so you are a Dane ?
I am in awe of your mastery of English, congrats

back on topic
K2, why not write an 'update/clarification' article - could dispel some more myths

be cool
 
K2 said:

SysCrusher
True, Aluminium-alloy 6061 is a little more corrosion resistant than pure aluminium, but for different reasons than the one you mentioned. The alloying elements help stabilize the protective oxide layer, because they influence the crystal structure.

True, I've looked into it further and that is what I discovered. You and Athlonnerd are correct.

K2 said:

You mention that copper ions are responsible for the corrosion, but where do they come from ?. If you put a copper block in distilled water, there are no or only trace amounts of copper ions in the water. Copper does not spontaneously decompose into ions and electrons, if you have read otherwise, I would be very interested in reading it, as it basically contradicts what I have learned.

This is how I understood it which I could be wrong but makes perfectly good sense to me - Then again I have no Phd. in science. The aluminium would be the anode in our case. Question to ask is. What would happen if we introduced zinc or magnesium in place of Aluminium? What would that interaction be between the two and why?

The perfect test would be to get three small pieces of aluminium, a steel screw, piece of copper, zinc and three cups of distilled water. Pair each piece of aluminium with zinc, steel screw and copper. Make sure they don't touch. See which ones corrode/pit then ask why.

It's been two weeks with my 6061 block and no pitting or corrosion. Altleast no pitting that is noticable to the eye which is what I'v been worried about. I even use tap water right now. It's only been two weeks though so I'll see what happens. It just might corrode.
 
BillA:

I am 100% Danish, but I try not to sound like one, it seems to work, thank you. MS Word helps a lot.

Pitting corrosion is related to impurities in aluminium, so the purer the aluminium, the less likely corrosion will be. But you can never get aluminium clean enough to avoid pitting, instead you must preserve the oxide layer, by avoiding chlorine.
Very pure aluminium is expensive, so you are better of with copper anyway, and copper has better thermal properties anyway. It is definitely not worth the effort, I simply use an aluminium radiator because it was dirt cheap.

I think I will write an article, to maybe clear a few things up and maybe dispel a myth or two, perhaps garnished with some experiments, as SysCrusher suggests. The problem is as always finding the time.

SysCrusher:

I am all in favour of more experiments, but I think the results obtained would have greater value if all metal parts are weighed before and after. Sometimes corrosion is in detectible with the naked eye, a properly calibrated scale will reveal all. Also you should have some kind of circulation in the liquid, to better mimic water cooling conditions. Unfortunately I don’t have access to a proper lab at the moment, but I might be able to persuade a friend to help.

If your local water supply is of good quality, you might not get any problems what so ever, but 2 weeks is a bit early to tell, I would keep an eye on it though.

Since87:

Sorry I didn’t comment on the thread you linked to earlier, somehow I missed that. Corrosion is very real in watercooling, but galvanic corrosion is extremely rare or nonexistent. It is impossible to pass judgment on the examples mentioned, based on a couple of pictures, but I think roscals pictures again show pitting, not galvanic corrosion. To be absolutely sure what the cause is, you need lots more information and pictures. It is like performing an autopsy without a body.
Thank you for your welcome by the way.
 
SysCrusher
since chlorine is the 'disinfectant' of choice here in the US (in almost all municipalities), despite it's virtual banning in Europe, you should keep an eye on your Al

and note that once the pitting has initiated it can be quite vexing to stop
takes a good inhibitor and some time to be effective
(my experience over several years with mixed-metals systems)

be cool
 
K2 said:
Since87:

Sorry I didn’t comment on the thread you linked to earlier, somehow I missed that. Corrosion is very real in watercooling, but galvanic corrosion is extremely rare or nonexistent. It is impossible to pass judgment on the examples mentioned, based on a couple of pictures, but I think roscals pictures again show pitting, not galvanic corrosion. To be absolutely sure what the cause is, you need lots more information and pictures. It is like performing an autopsy without a body.
Thank you for your welcome by the way.

I'm glad to have you educating us. I wish it had come a little sooner because I just sold a nice aluminium heatercore a week ago out of paranoia combined with some stuff BillA had written about problems with Water Wetter. Oh well. I'll have more peace of mind with a copper/brass heatercore anyway.

It does seem to me, that I've observed a significant correlation between mixed metal systems (with electrical connection), and rapid corrosion, but that may be observer bias on my part. (Far from being a scientific analysis.)

I hope you will write an article. You seem to have a better understanding of the underlying issues than anyone else I've seen post on this subject.
 
Back