227

I have read that distilled water doesn't conduct electricity. This, in other words, means that we can submerge electronic devices like PCs/laptops in it and run them without any problem. I haven't seen much information about this on the internet, but it should be possible.

So, can you really run a PC in distilled water? I don't know if you can, but I think if you could, it would start rusting/corroding in a few days. ;)

  • 83
    I don't think that this is feasible in real life. It is correct that distilled water is an insulator, but as soon as contaminants are introduced into it (e.g. from the tiniest amounts of dirt on the boards, fingerprints etc.), it loses its insulating properties. – Nassbirne Jun 10 '16 at 13:22
  • 28
    What you really want for this trick is de-ionized water. But there's no way for you, outside of a lab environment, to keep a tank of water from picking up contaminants over time that will give you trouble. If you want fluid cooling, try mineral-oil. It's got it's own issues, but it's far less troublesome than water for the average user. – Michael Kohne Jun 10 '16 at 13:56
  • 17
    So, the idea is to immerse a motherboard in the Universal Solvent. The solvent will quickly make its own contaminants. water.usgs.gov/edu/solvent.html – Don Branson Jun 10 '16 at 17:39
  • 20
    @Ramhound: what you're saying doesn't make sense. Water has a pretty vast specific heat capacity, hence it is a much better coolant than air (also a lot better than oil), precisely because it does not heat up quickly. It fact this also means that it does not cool down quickly – the heat eventually needs to go somewhere, this is just conservation of energy. But a large volume of water has much more surface area to dissipate heat than a few chips. — None of this has anything to do with whether the water is pure/distilled or has electrolytes in it. – leftaroundabout Jun 11 '16 at 11:16
  • 10
    @JavaLatte Pure (de-ionized) water has a resistance of 18 MΩ, which makes it a pretty good electrical insulator. If you could keep contaminants out, it would make an ideal coolant (others have mentioned the high specific heat of water). On the other hand, the wires would probably (very slowly) start undergoing an electrolytic reaction of some kind just due to the voltage applied, which would put more ions in the water, and drastically increase the conductivity. – Riet Jun 12 '16 at 13:37
365

I've done it. Don't do it.

I set up a computer in an acrylic case with good quality distilled water and a cheap motherboard as a test, with heatsinks only (no fans/moving parts). I cleaned the inside of the case with isopropyl alcohol, thinking that would remove any existing contaminants.

Within a day or two, I noticed that all the contacts/metal parts on the board began to rust. Even the stainless steel on the case of the SSD had begun to rust. Another day later, the motherboard died. When I removed the motherboard, being the first time anything physically removed (no fans), a huge cloud of rust particles came off and turned the water a lovely brown color.

Stick with something that metal parts can be friends with, like mineral oil.

  • 27
    And all that rust was due to electrolysis of water I suppose. – Ruslan Jun 11 '16 at 10:37
  • 35
    @Ruslan It is not to due to the electrolysis of water. It is due to the fact the water is "begging" for minerals to be dissolved. It is highly undersaturated, that is why it tends to be corrosive. – cinico Jun 11 '16 at 18:23
  • 14
    Rust requires water, oxygen and heat. Water and heat are available, and I guess the water was saturated with oxygen from the start or after handling it. Would be intersting to see in a lab enviroment if without oxygen the whole system would work better. Though I think in the long run the lack of real cooling and heat dissipation could be a problem, depending on the volume of the water and the purpose of the computer. – HopefullyHelpful Jun 11 '16 at 18:45
  • 14
    Heh, dissolved oxygen is not your friend, first you deionize your distilled water, then you introduce it to a nitrogen or argon environment. The whole exercise is a total waste of time for other reasons mentioned by others. – Fiasco Labs Jun 12 '16 at 2:49
  • 82
    So, you're saying he should write his code in Rust? – Mehrdad Jun 12 '16 at 4:19
109

Yes it is. Running a computer in distilled water is no issue.
However, keeping the water distilled is near impossible.

As soon as contaminants pollute the water even in very small amounts, the water will begin to corrode and given enough ionic contaminants, the water will stop being an insulator and become a very good conductor.

This kills the computer.

Now various people will say different things with regards to the amount of time it takes for the water to become contaminated enough to cause problems but in almost all cases it is within weeks in sealed environments, days in open.

Mineral oil is a far better alternative for a submerged build.

  • 2
    Mineral oil has it's own problems as has been noted by another commenter. Please do further reading on pro's and con's before carrying out anything of this nature. – Ctrl-alt-dlt Jun 10 '16 at 13:58
  • 3
    Even if the water is kept unconatmined with anything, afaik H2O + O2 + Metal = Rust. Rust probably has other electrical properties than the original Metal, which might cause the motherboard to not be part of the electrical closed circuit anymore, as it's connection is severed by a wall of rust. – HopefullyHelpful Jun 11 '16 at 18:49
  • 2
    @HopefullyHelpful H2O + O2 + Metal = Rust? Wrong! Only metals containing IRON will rust. Other metals will corrode too, but it won't even look like brown rust... Copper turns green, and aluminum-oxide is white... – svin83 Jun 13 '16 at 7:21
  • 1
    Oxygen isn't needed for it to rust. The solder on the anode can react with water. – v7d8dpo4 Jun 13 '16 at 9:21
  • 5
    I love how you say "in almost all cases" which also serves as "computer cases" here :) – Konerak Jun 14 '16 at 13:39
42

I would be highly surprised if it actually worked, even for a second. Motherboards have some pretty high frequencies, and the PCB routing is intricately designed to minimize capacitance so that they can actually carry these signals.

Changing the fluid that is around the board from air (dielectric constant = 1.00059) to water (80.4) is likely to introduce a lot of capacitances that weren't designed for and would be way out of tolerance, especially for channels like CPU to RAM. The additional capacitance just wouldn't allow the signal to switch fast enough to be able to reliably transmit the data. By the way, mineral oil has a dielectric constant of 2.1, so much less capacitancy than water, and some people have had success with submersion in that.

If you would be doing this so that you can overclock everything, then the higher dielectric constant works against that by reducing the maximum frequency that the board can operate at.

The Cray computers didn't have nearly the same challenges to being submerged, since the highest fundamental frequency signal on the board was 125MHz, and modern machines potentially have ~4000MHz signals, with common RAM being just below 2000MHz, with harmonics extending to >5x the fundamentals to form the waveform accurately.

I agree with the others here that have noted that metals are slightly soluble in water (especially copper), so the water would start to become conductive immediately. Voltage differences would also cause electrolysis through the water and H2 + O2 would be produced, as well as forcing ions into aqueous solution.

  • 1
    Well, at the risk of posting a link that may become outdated: electronics.stackexchange.com/questions/56574/… Wikipedia: en.wikipedia.org/wiki/Parasitic_capacitance – Keith Procter Jun 10 '16 at 20:13
  • 1
    Fluorinert has a dielectric constant of 1.9, so supercomputers run in something like that wouldn't have the capacitance issues that they woudl in water. Also this paper demonstrates that the capacitance properties of air vary a lot with humditiy. – Chris H Jun 13 '16 at 10:43
  • Mismatched lines can already be tremendous trouble at 125Mhz :) Some of the capacitive (and resistive) effects will be reduced, though, by the lacquer that is on most traces of a modern motherboard; extra capacitance on bare chip pins will of course still be a problem. – rackandboneman Jun 15 '16 at 13:38
14

I cannot speak to the use of water but a liquid cooling system was implemented years ago using fluorinert. This was done on the cray 2 and 3 I believe. The following snippet can be found on wikipedia. I did have the opportunity to see the cray-3 running in a tank of fluorinert completely submerged in liquid much like a fish tank.

The cards were packed right on top of each other, so the resulting stack was only about 3 inches high. With this sort of density there was no way any conventional air-cooled system would work; there was too little room for air to flow between the ICs. Instead the system would be immersed in a tank of a new inert liquid from 3M, Fluorinert. The cooling liquid was forced sideways through the modules under pressure, and the flow rate was roughly one inch per second. The heated liquid was cooled using chilled water heat exchangers and returned to the main tank. Work on the new design started in earnest in 1982, several years after the original start date.

  • 2
    If you want to do this with your own PC, it is possible. But I'd advise you to find a source of recycled fluorinert -- the stuff is extremely expensive. – Jules Jun 10 '16 at 19:32
  • 7
    Fluorinert is also useful if you want to send Ed Harris to the bottom of the ocean. – hobbs Jun 12 '16 at 23:43
  • 6
    Something a little more modern is Novec, which is a 3M product. It's used for everything from fire supression (where water would damage equipment) to cleaning electronic equipment. Unlike oil, which can provide problems if you need to change hardware, it won't stick to your hardware. It's used in some CPU coolers ('water' cooling) and I've seen a display at one of their sites where a normal mobile phone was submerged and fully operational, allowing you to send texts or call it. – Baldrickk Jun 13 '16 at 15:34
  • Some models of the ETA-10 submerged the CPU in liquid nitrogen. Nitrogen is easy to obtain. You just need a good way to cool it down enough – Theodore Norvell Jun 18 '16 at 16:21
9

It would seem that pure water would not cause any electrical problems given its insulative properties, and it's further suggested that you would want deionized water, but the problems that arise are only partly due to the introduction of contaminants (e.g. minerals, salts, metals, etc.). Even if you could guarantee that no contaminants entered the water, problems are inevitable on account of the autoionization of water. Neutral water does not remain neutral.

  • 2
    having un-insulated electrical circuitry running in water will ionize it quickly too... – svin83 Jun 13 '16 at 7:24
  • 1
    @svin83 It sure will. You essentially have a bunch of cathodes and anodes in the water. – benJephunneh Jun 13 '16 at 14:54
  • @svin83: Only if it exceeds 1.23V. DDR4 runs at 1.2V and is safe, but PCI-e causes problems. – MSalters Jun 16 '16 at 10:43
  • @MSalters: Care to elaborate on why it has to exceed 1.23V to become a problem? – svin83 Jun 16 '16 at 12:53
  • 1
    @svin83: 1.23 Volt is the voltage needed to electrolyze water into hydrogen and oxygen. It means you don't just have auto-ionization from 2*H2O => H3O+ & OH-, but you are also getting all the ions from the electrolysis process. – MSalters Jun 16 '16 at 14:13
2

As the water (in conjunction with oxygen which is always in water, taken from the air to some equilibrium) would corrode metal parts, you have to prevent the metal parts come in direct contact with the water.

This can be done by painting of the components in some water resistant finish. There are several coatings out there exactly for this purpose, protecting electric components from water. Although this paints are meant for occasional dew, some of them work quite well for total submergence.

You just have to made sure your finish doesn't break contacts that are needed (just spray paint after connecting all plugs needed) and doesn't stop cooling (eg. keep the paint off the CPU heatspreader or sand it to a very thin layer there).

While some special praised paints doesn't seem to provide a long term protection (see here: http://hackaday.com/2013/12/26/neverwet-on-electronics/ ), more simple plastic sprays or expoxy based resin paints may do if the layer is thick enough.

-1
H2O   

does not conduct electricity, however distilled water is more like

H2O  <-> H20 + H + OH

Actually the % of the ions is really low

just 10^-7

So every about 10.000.000 molecules of water you have also a H and a OH ion. (If I remember correct my studies over pH, in case I'm wrong just let me know I'll refresh some book or look at wikipedia)

but enough to cause troubles in the long/short run (depending on intensity of current and magnetic fields)

And here you need just a minimal differential of potential to cause water be subject to electrolisis, hence the ions will be pulled out of water thanks to the magnetic field and will react with metal parts.

So actually you have ions inside water that can still carry charge (and so electricity, even if low currents), and despite that any magnetic field, even minimal will cause ions to separate from water and hence attack any metal part (beacause also parts in different metals act as catode-anode)

In reality water is corrosive for metals even without currents (well technically the metals will create a current even if not plugged into a power source), but current can accelerate/mitigate the corrosion (of course since computer parts are not designed for that, it is likely that a computer part would provide the exact current to counter the corrosion and hence will corrode.

  • 1
    The positive ion is actually H3O+; no free hydrogen ions in water. That makes a big difference here because the H3O+ ion is much bigger and heavier than a H+ ion, and a much worse conductor. And the pH=7.0 is for pure water at room temperature; heating causes quite a bit of extra ionization. – MSalters Jun 16 '16 at 10:46
  • Thanks I suspected I was remembering that in a wrong way (6 years from last chemistry exam XD) – GameDeveloper Jun 16 '16 at 10:54
  • All those un-insulated conductors will become anodes and cathodes as soon as the board is in contact with water, so the ionization will only accellerate no matter how pure your water is... There will always be impurities.... The Motherboard and all other components ARE impurities too... Unless you clean them all thoroughly first... Thermal paste, residue from the manufacturing, fingerprints, dust etc. will also contaminate the water. and different metals in water, several of them conducting electricity... NO DICE. – svin83 Jun 16 '16 at 13:00
  • There are also trace amount of deuterium and tritium in water. Any chance that the rig could turn into an atomic bomb if it ran long enough? :-) – fixer1234 Jun 16 '16 at 19:02

protected by Community Jun 12 '16 at 21:42

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.