Copper Sulphate PentaHydrate as a coolant solution

[Cathar]

Been playing in a small way with copper sulphate pentahydate as a cooling liquid solution.

In particular relation to the thermal properties of the substance, it has 4x the thermal conductivity of water. Have been trying to dig up some information on its kinematic viscosity, but am not having much luck. If it's as viscous as water is, or even if it's 5x more viscous, it's still a better solution to use than pure water in pure convectional liquid sense.

Eats away at nickel though pretty savagely. Seems to have no affect on 316 stainless steel. It's unreactive with copper and silver. Slight reaction with brass. Unreactive with most plastics.

Anyone got anything more to contribute. It has about 1/3rd the thermal capacity of water per volume, but any flow rate above 4LPM effectively negates that as a significant issue.

pH is between 3-3.8 depending on the concentration of its dilution.

Anyone else want to contribute?

[pHaestus]

It's a nice color

Don't put it in contact with Zinc; it'll immediately displace it in whatever solid phase. If you think about how oxidation of copper proceeds:

Cu(s) --> Cu2+(aq) + 2e-

Then it makes sense that you should slow the oxidation of copper blocks with copper sulfate in solution due to Le Chatelier's principle. In terms of long term use in a cooling loop, my concern would be that you should never use it with water wetter or antifreeze (copper hydroxide will begin to precipitate above pH 5 or so). Sulfate is a tolerable electron acceptor but as long as your loop is free of sulfate reducing bacteria then I see no huge problems with microbially induced sulfides. Any idea about the wetting angle of a concentrated copper sulfate solution? That would be the only other thing that could screw cooling up I guess...

[gruntledweasel]

As far as I can tell, this document lists details of a commercial product that is 13% copper sulphate pentahydrate by weight. pH is 2.4 (more concentrated than what you were looking at?). Viscocity is listed as 15 cps @ 15 degrees (Celsius?). I don't know what those units are, but I assume they're what you're looking for?

click me!

The stuff's reasonably environmentally friendly (used as fertilizer) and google tells me it's a good biocide.

edit: Only off by ten to the third on a number I mentioned.
edit, the sequel: it's cheap on ebay, if anyone wants some.

[Cathar]

Originally Posted by gruntledweasel

Viscocity is listed as 15 cps @ 15 degrees (Celsius?). I don't know what those units are, but I assume they're what you're looking for?

cps = CentiPoises

That's absolute viscosity, but it'll do.

Crap.

Water has an absolute viscosity of 1.12cps @ 15 degrees, making Copper Sulphate over 10 times more viscous than water.

So long as the solution is less than 10 times as viscous, it offers some advantage due to its higher thermal conductivity, but above 10x more viscous it holds the boundary layer enough to outweigh any advantage conferred by its higher conductivity.

At least that's for a couple of impingement cases I was working with.

So, no-go then...

[gruntledweasel]

Originally Posted by Cathar

Water has an absolute viscosity of 1.12cps @ 15 degrees, making Copper Sulphate over 10 times more viscous than water.

Um, maybe I'm missing something here, but that wasn't a pure copper sulphate solution I linked to (the pure stuff isn't helpful, as the melting point of the salt is over 100 degrees C). It was a solution of the salt dissolved in water. So a solution half as concentrated would fall within your parameters, right? Or does viscosity not work like that?

[Crapgame]

Originally Posted by Cathar

Been playing in a small way with copper sulphate pentahydate as a cooling liquid solution.

In particular relation to the thermal properties of the substance, it has 4x the thermal conductivity of water. Have been trying to dig up some information on its kinematic viscosity, but am not having much luck. If it's as viscous as water is, or even if it's 5x more viscous, it's still a better solution to use than pure water in pure convectional liquid sense.

Eats away at nickel though pretty savagely. Seems to have no affect on 316 stainless steel. It's unreactive with copper and silver. Slight reaction with brass. Unreactive with most plastics.

Anyone got anything more to contribute. It has about 1/3rd the thermal capacity of water per volume, but any flow rate above 4LPM effectively negates that as a significant issue.

pH is between 3-3.8 depending on the concentration of its dilution.

Anyone else want to contribute?

You shouldn't need any algicide in the loop using it. Copper sulafte pentahydrate mixed with water is sprayed on farm ponds to kill algae.

As far as water wetter, there are plenty of surfactants out there that work in acids so keeping the ph down below 5 while using a wetter should be easy. I used to work in a chemical plant where we made the solutions used to silver plate and copper plate glass for mirrors and worked with copper sulfate pentahydrate every day. You might want to check into the chemical resistance of your pump and tubing before you try it. For a mirror line we used polyethylene tubing and pretty expensive diaphragm pumps designed specifically for chemicals (but our solutions were very concentrated also, ph of1-1.5 on the acid solutions, a ph of 13-14 on the base solutions).

You could look into the thermal properties of silver nitrate solutions also, but since it's photoreactive it's pretty nasty to work with. I had thought about chemically silver plating waterblocks for looks since it's pretty easy to do versus electroplating ( I silver-plated a dragonfly once when I worked there). I doubt it would be of any thermal benefit but it looks nice.

[Butcher]

Tubing should be fine if it's Tygon or similar as that sort of lab-grade tubing is designed to carry acidic chemicals.
Pump may be more of a problem.
Also, what's the specific heat capacity of copper sulphate pentahydrate?

[myv65]

Cathar,

Viscosity is far more important than the thermal capacity or conductance of the fluid. Viscosity affects both the boundary layer as well as the flowrate/delta_P (power vs flow) relationship. You generally are looking at relatively high flow rates, so your delta_T across the block is negligible, hence thermal capacitance not having much impact. Conductivity will play a role in how efficient/inefficient the transfer through the boundary layer is, but this is not as major a player as the viscosity. Sure, you'll transfer heat through the layer more effectively, but the layer will be thicker. Factor in the added pump juice needed to maintain flow and it seems like a losing proposition to me.

In very dilute concentration is makes a wonderful algaecide. I wouldn't consider it for much more.

[|kbn|]

I would have thought thermal conductivity for the coolant is irrelevant seeing as the idea is that the pump is suppost to move the coolant (and the heat it carries) instead of moving the heat from to coolant using conduction.

Heres my post from anothe foum on the same question (about minerals in water helping performance)

I doubt minerals in water help it remove heat, I wouldnt mind seeing some data supporting that. for coolant we dont really care about conduction, its specific heat we need....

It might be worth looking into, but not with copper. Water is only used because it has a high specific heat - the amount of heat needed to raise its temperature. IIRC copper has a lower specific heat (but a much higher thermal conduction).
Water is the best / most practical liquid for moving heat in pc cooling atm.
There might be better solids that can be used if made into very fine powder.

The powder would coat the insides and eventually decrease performance as more of it gets deposited leading to small areas of static coolant - the solid powder covering the waterblock/rad (which would be where the water moves slowest - res or rad, so it would need to be cleaned after a while.

Also think about how it would affect flow through waterblocks.

considering the ratio of powder to water, and the maximum amount that you could practicly use, I dont think this is worth anymore thinking about....
.. unless theres a material about 100000x better than water for specific heat - I doubt it.

[myv65]

Conductivity matters primarily for the boundary layer. In the boundary layer, there is very little fluid motion. Without fluid motion, heat can only be transferred via conduction. Practically speaking, the boundary layer behaves much the same as a solid, hence conductivity does matter.

[Boli]

copper sulphate pentahydate?

Is this copper suphate solution that you use in chemistry to grow crystals, if so when it starts to evaporate cystals will most likely start to form possibly becoming a hazard to cascade cups for example.

on the up side though it is poisonous and nothing will grown in it so you'll see a nice clear blue all year round without cause to add additives .

Change the molarity of the solution (1/4 water maybe) to decrease its viscosity and maybe it will be more usable and still retains a higher thermal conductivity than water.

I think the biggest problem you'll have to overcome though is crystal formation. that and ANY currant that passes thorough it will cause problems through corrosion, and your block will just become slowly eaten away... though if you ever want to copper plate something you can just dip it into your res on a wire attached to a battery, and the other end touching your waterblock

[|kbn|]

hmm I understand now why this could be a good idea. Another question then: how will using a copper sulphate pentahydate solution affect the specific heat and conductiviy and viscosity for different ratios.

it reacts with brass too I would need two new heatercores

[Cathar]

Thanks for the comments Dave (myv65).

Am aware of the effects of viscosity on jet impingement efficiency. Was talking primarily here about a fairly narrow range of impingement scenarios though where we aren't moving too far away from the edges of the jet (r = 1.0-2.0d from middle). Viscosity in this region under the primary influence of the incoming jet is of reduced overall impact when the jet nozzle is brought down to about 1.5-2.5d distance due to the pressure of the jet stripping away the boundary layer fairly effectively in that small region.

What we're left with is the thermal conductivity of the liquid then being of prime importance. As I noted, bringing the flow rates up above 4LPM or so effectively nullifies any loss due to the fluid's lower capacitance (as you also pointed out).

What I was seeing was that within that fairly tightly defined scenario, a thermal conductivity of 4x of water (which CuSO4.5H2O has) places it better than water for heat removal even if it had a viscosity of as much a 10x higher than water. Above 10x higher viscosity and water wins again.

From what I've been able to find out, it seems that CSP has a viscosity of around 12x that of water, so basically we're back at square one - water is still better.

Boli, yes. that's the stuff. It will only form crystals if the amount of water in it gets below the saturation point of the substance, where it will dry out into a crystallised form.

Lowering the concentration is of no real use. Yes, that lowers the viscosity, but it also lowers the thermal conductivity at about the same rate, so that effectively kicks out any gain from even doing that...

[myv65]

I don't know, Cathar. 10X viscosity means a heckuva lot more umph from the pump. I remain convinced that any gain in block efficiency goes bye-bye due to overall thermal energy dumped into the system. Yeah, a really big radiator will still get you back to the same temperature regardless of the pump power, but it gets outside the range of "practical".

From the standpoint of delving into the deep end, by all means have at the experiment. I'll be shocked if it represents any sort of practical application. Then again, when has original research ever had much to do with being practical? Best luck on the experimenting and keep us posted.

[Cathar]

Originally Posted by myv65

10X viscosity means a heckuva lot more umph from the pump.

True, I was just looking at a pure impingement efficiency thing, without considering the pump.

Had CSP had the same, or even just 2x the viscosity, of water, it would've made for a viable coolant.

Still, at lower concentrations it will act as a decent biocide, without some of the thermal unglosities of various additives, so it still remains as a moderately decent additive to stop the nasties and not kill convectional performance as a result.

[UberBlue]

I'm probably out of my depth again, but we used that stuff on the farm to cure hoof rot in sheep.

I remember it was a vibrant blue solid that dissolved readily in water. And the water didn't seem noticeably more viscous.

You can also use it to put a quick coat of copper on steel.

EDIT: It just dawned on me what the signifigance of the "penta hydrate" is. Close but no cigar.

[Boli]

if the viscosity becomes such a problem then making the copper sulphate pentahydate an additive that you put in to:

1: colour
2: increase the thermal conductivity (slightly but better than nothing)
3: kill anythign in there

Is a viable alternative to the higher concentrations. Really though we could stand and argue this for months testing is the only real way to go. i.e. see how well the pumps perform under long periods of stress.

[bigben2k]

I remember spec'ing out my LGPC pump for different viscosities, and it was surprisingly untolerable. Where the water's viscosity is 1 cps (right?), the pump could handle up to 2 cps, which is nowhere near any light oil.