2

I've read conflicting descriptions of the corrosiveness (or lack thereof) of very pure water. For a precision water bath, it's stated that one should fill the bath with distilled water rather than deionized of a certain purity as it would be corrosive:

5.2 Filling Bath
NOTE: DISTILLED WATER IS RECOMMENDED; IT WILL NOT CORRODE THE BATH CHAMBER AND IT REDUCES THE NEED FOR FREQUENT BATH CLEANING.

DO NOT USE 18 MEG DEIONIZED WATER.
NOTE: If this is the only source of treated water available; mix 50% with regular tap water.

Why is this? I'm struggling to find a specific mechanism by which purer water would corrode the bath.

Buck Thorn
  • 21,621
  • 5
  • 38
  • 89
user126124
  • 31
  • 1
  • 3
  • 1
    Because there is a strong entropic driving force to suck impurities out of everywhere. – Jon Custer Aug 03 '22 at 02:00
  • @Jon Custer When it comes to how the manual describes DI water as more corrosive than distilled, how exactly is that occurring? Doesn't there need to be a conductive medium to accelerate the process? Distilled is highly pure as well, so I'm not grasping the difference. – user126124 Aug 03 '22 at 16:46
  • @Mithoron Somewhat, but not much more an answer than "the manual is just wrong" which seems to be what I'm picking up from reading similar questions here and elsewhere. I'd really like a clear mechanistic explanation to support this claim in the manual, but perhaps it is just incorrect. – user126124 Aug 03 '22 at 20:55
  • What about manual? That they say not to use "18 MEG DEIONIZED WATER"? Well I don't see any suggest about it being "corrosive". I wouldn't suggest using it because it's too expensive. – Mithoron Aug 03 '22 at 21:56

2 Answers2

2

You can stop struggling. Your chemical instincts are correct; the purer water is, the less corrosive.

The main factor in the corrosivity of water solutions is the ion content, the ability of the water to carry a current in galvanic reactions. Lower ions, less current, less corrosion. The second factor is the presence of dissolved ions etc. to participate in various and sundry reactions. The less of these the fewer possible reactions. It is best to use water of the highest quality available in water baths, humidifiers, drip coffee makers, CPAP devices, plating baths, even laboratory dishwashers, and definitely as a chemical reactant or solvent. Printed circuit board manufacturing uses only the highest quality DI water in any aqueous process [and they probably wish they could get rid of the pesky 10^-7M H3O+ and OH-]. [Which they do by inventing these super nonionic detergents and no residue fluxes and, once upon a time, freon rinses.]

The increased corrosiveness of pure[r] water is a myth that I think originated with the water supply companies that found that certain plumbing structures literally fell apart when high calcium water was replaced by softened water. The pipes were partially held together by buildup of CaCO3 deposits and the common ion effect prevented dissolution of the deposits. Remove Ca++ from the water the deposits dissolved and everything leaked or even fell apart. This was probably part of the problem in Flint, MI. A change in water disturbed decades of protective coating in the pipes; the problem was deeper than removal of a coating.

jimchmst
  • 2,170
  • 4
  • 9
  • You are missing the fact that pure water will happily suck atoms out of the piping. Not electrochemical corrosion, just a question of trying to come into equilibrium… – Jon Custer Aug 03 '22 at 17:02
  • 4
    @JonCuster Oh, no, not you too. Only effect of such "sucking" is that water gradually stops qualify as distilled or deionised. And only equilibrium would be if metal was oxidised completely - saltwater is incomparably better in achieving that. – Mithoron Aug 03 '22 at 17:58
  • 1
    Jon, Look up the oxidation potentials of pure water compared to standard potentials to see how efficient pure water is in "sucking out ions". Water of different ion content has different effects on mineral deposits. Metals become ions by electrochemical corrosion mechanisms and an essential part of corrosion is an ionic pathway between anode and cathode no matter how close the two might be [at least down to the actual molecular level then there might be an explosion if enough reactants are present.] – jimchmst Aug 03 '22 at 19:00
  • @jimchmst - consider the Gibbs free energy when pure water is in contact with, well, anything. The entropic term far outweighs the enthalpy term until some impurities are introduced into the water. In a semiconductor fab, all DI plumbing is plastic for a reason - even ppb of many metals are a problem and it only takes one small section of metal plumbing to provide that. Reconsider your electrochemical argument when the system is that far from steady state much less equilibrium. – Jon Custer Aug 04 '22 at 01:09
  • 1
    @JonCuster Do the calculation. Don't just assume the claim. In every practical way, pure water is not corrosive. Yes, contact with metal surfaces will make it less pure but the effect is vastly smaller than the effects seen with impure water where the impurities often increase the surface losses by electrochemical processes. So the surface losses to pure water are vastly lower than to dirty water. – matt_black Aug 04 '22 at 10:50
  • Jon I have chemistry on my side and also practical experience I have worked in PCB manufacture, electroplating, water purification and chemical analysis and have never found that using better grade water created problems. In almost everycase there were subtle to sometimes dramatic improvements in processes when we improved water quality. Things as simple as using RO water to make reconstituted orange juice to using 18Meg water to wash components in PCB manufacture. The only problem with superpure water is that attention must be paid to keeping it super pure. – jimchmst Aug 06 '22 at 02:31
2

NASA published a 23-paged reported in the 60s on the corrosion of metals in deionized water at 38 $^o$C. The report is titled "Corrosion of Metals in Deionized Water at 38° C (100° F)" By Barbara Alice Johnson, 1969.

As one can see, we cannot make sweeping statements about corrosion in deionized water. Some metals will corrode and some will not in deionized water. Long time ago, someone told a story of detecting measurable amounts molybdenum in human blood. They were using very sensitive analytical technique and it puzzled everyone as to why Mo is being detected. It turned out that the metallic syringe needle leached a few atoms of Mo in a very short period of drawing blood. Yes, deionized or distilled water will leach some metallic ions but corrosion means measurable metal loss to the extent that it causes economic damage.

Thus, corrosiveness distilled vs. deionized water is like a popular urban myth. What the manufacturer probably want is some level of slow scaling in the bath to "protect" metallic heaters.

enter image description here

AChem
  • 40,037
  • 2
  • 63
  • 126
  • Corrosion is corrosion If you wait for economic loss you have to wait for the bridge to collapse. Think of all the money that would have been saved by not painting the Golden Gate Bridge. A chemical engineer's job is to understand these things. Talk to Delorean car owners for an example of poor chemical planning. – jimchmst Aug 06 '22 at 19:31