Sciencemadness Discussion Board

Liquid with high thermal, (very) low electrical conductivity?

Vinatta - 23-7-2006 at 00:23

I'm looking for a liquid with high thermal conductivity and extremely low electrical conductivity with a reasonable cost. Non-toxic, stable, and ease of working with is a must. I'd like to undertake a project similar to the one here: Liquid Cooling for PSU

Chemistry, while an interest of mine, is not at all my strong point. I hear some transformers are submerged in highly refined mineral oil, but I would not know the difference between this transformer oil and off-the-shelf mineral oil.

How electrically conductive is OTC mineral oil? What would I want to research/look for in selecting a high-grade refined mineral oil?

I really appreciate your advice!

unionised - 23-7-2006 at 00:31

The baby oil that they use on that site seems like quite a good idea. As long as you are not working with high voltages then practically any oil will be a good enough insulator. High voltage insulation is an art in itself and needs a fair bit of study before building stuff. I have used cooking oil at 20KV, but I can't exactly reccommend it as safe. Medical grade liquid parafin is also used as an insulator by HV DIYers.

not_important - 23-7-2006 at 00:50

Go for unscented oils, then you're dealing pretty much with pure alkane hydrocarbons. Not very reactive, decent insulator.

HOWEVER, it may act as a slight solvent and slowly soften plastics, including the PC board. Transformers are one thing, copper and steel and insulation, a computer is another. Military gear would be submerged in fluorocarbon liquids, which can be very poor solvents; you don't want to know the cost.

IrC - 23-7-2006 at 01:01

Shell Diala-X is the most used transformer oil. If very high temperatures are expected then silicone oil is a good choice but more expensive. Unless you need many gallons cost is not that great.

Vinatta - 23-7-2006 at 01:37

Very high-bound temperature limit would be 80 Celcius.

I definately need a very inert liquid. Having any plastics, alloys, non-metals, or semiconductors dissolve into the liquid could be very bad.

I'd love to spray the boards with teflon, but it's a great thermal insulator, which I'm obviously trying to avoid. Would a very thin coat work and still provide heat transfer? Or is there another chemical I could look at?

Silicone Oil looks promising. Any concerns I should have with it?

(I really dont have any idea what to expect in terms of unwanted reactions vetween the circuit boards and coolant solution X)

P.S. Thanks for the tips in getting me started. After searching with "silicone oil" and "Shell Diala-X" I ran across this thread http://www.hardforum.com/showt...1&pp=20. which details one or two real-world experiments with various coolants. I don't know how to interpret some of the information though. Such as, if an organic coolant oil starts breaking down, what properties would it gain/lose? (i.e. Diala-X, luminol TR-I)

[Edited on 23-7-2006 by Vinatta]

DeAdFX - 23-7-2006 at 05:54

I was thinking of SF6 (sulfur hexafluoride). The chemical is a gas and fits your requirements of high thermal conductivity and low electrical conductivity. I do not know where it can be purchased from or its cost.

ethan_c - 23-7-2006 at 06:58

Quote:
Originally posted by DeAdFX
I was thinking of SF6 (sulfur hexafluoride). The chemical is a gas and fits your requirements of high thermal conductivity and low electrical conductivity. I do not know where it can be purchased from or its cost.


It's certainly not available OTC to my knowledge- also, when it's used in high-voltage operations, as a heat conductor and electrical insulator, it is always very pressurized (just on the edge of liquid state, I think) to increase both of these properties- it is still a very good electrical insulator while not under pressure, but I would imagine that heat transfer would suffer significantly.

woelen - 23-7-2006 at 10:13

If you can get ahold of some calcium metal (sometimes on eBay) or sodium metal, then the standard paraffin oil is a good option. You need to put some calcium metal in the oil for a few days in a closed container (with some air trapped in as well). You'll see that very slowly still some hydrogen is evolved at the pieces of metal, due to traces of water and possibly also hydroxy-containing organics. These all are trapped by the metal, the metal being covered by a very thin hydroxide (possibly also some alkoxide) layer. After this treatment, the oil is absolutely free of water, and then it can serve as a non-toxic, easy to use (although sometimes very messy) excellent isolator.

YT2095 - 23-7-2006 at 10:33

yup agreed, Calcium is the ideal metal of choice here, group 2 metals are largely insoluble also, and so you may be sure that your material will be safe for electricity (the most you`ll encounter in a Switchmode PSU will be about 400v DC anyway).
Sr and Ba metal`s expensive and Mg`s not reactive enough.
my Vote is Calcium also :)

12AX7 - 23-7-2006 at 11:09

If you tack on coolers to the heatsinks, water is fine, as long as you ground the tubing at the other end with a metal fitting.

I'm going to use water cooling for my induction heater, which I can assure you runs much higher voltage and frequency than is present on those heatsinks (which should be grounded to the power supply's chassis anyway, incidentially). For three feet of about 1/4" tubing, I calculated a resistance of over 100kohms. Big SWAG, but in the ballpark for hard water.

If you must submerge the whole thing, oil is definetly the way to go. If you want to clean it good first, you might soak the thing with ethyl or isopropyl alcohol (denatured or rubbing alcohols) and give it a shake to loosen things. Or even water. As long as you dry it out, it's fine.

Tim

Organikum - 23-7-2006 at 11:29

Ultrapure water?

Regarding water

neutrino - 23-7-2006 at 11:59

You may want to take a look at this site.
Quote:
Common sense dictates that submerging your high-end PC in cooking oil is not a good idea. But, of course, engineering feats and science breakthroughs were made possible by those who dared to explore the realms of the non-conventional. Members of the Munich-based THG lab are only too happy to confirm this fact. And not only did we find that our AMD Athlon FX-55 and GeForce 6800 Ultra equipped system didn't short out when we filled the sealed shut PC case with cooking oil - but the non-conductive properties of the liquid coupled created a totally cool and quiet high-end PC, devoid of the noise pollution of fans. The PC case - or should we say tank - also offered a new and novel way to display and show off your PC components.


These guys put a computer in a tank, filled it with ordinary cooking oil, and it worked. They also tried very pure water but they kept getting shorts.
Quote:
After the failed trial runs with de-ionized water, we quickly switched to oil. In contrast to water, this liquid has the crucial advantage that far fewer free ions (dangerous leak currents) are present. In the trials with water it had been determined that in the case of local shorts the components did not suffer any permanent damage. In the final analysis, the danger of losing $2,000 worth of hardware within a few minutes was thus averted.


I don’t see the need for fancy alkaline earth metals for drying oil. These guys got it to work with just ordinary untreated cooking oil.

To be on the safe side, I’d heat the oil on a stove to boil out any volatiles, but anything more than that really shouldn’t be necessary.

There was also a brief note on dissolving components.
Quote:
The aggressiveness of individual types of oil (above all those with a high fatty acid content) when it comes to plastics should not be taken lightly. For simplicity's sake, we decided on vegetable oil - but we recommend motor oil.


This seems like it would be hard to do in the long run.

[Edited on 23-7-2006 by neutrino]

Nicodem - 23-7-2006 at 12:20

Not to mention that vegetable oil gets oxidized by air oxygen making a mess out of it enough time given. It can even be attacked by bacteria and molds though I believe a PC has a shorter life span than any such biodegradability of the oil in question. Simple paraffin oil would probably be the best alternative to the "recommended" motor oil as it is perfectly clear of any additives that might cause degradation of the circuit or could decompose to conductive materials. It is also somewhat less viscous than vegetable oil making it transfer heat faster by convection.

12AX7 - 23-7-2006 at 19:29

Water will never work because it has a very high dielectric constant -- changing the precisely adjusted impedances of the board (200MHz+ is not to be taken lightly!), and also, no matter how clean the board and pure the water, ions will always dissolve and lead to malfunction.

Mineral oil is about the best you can do, aside from potential plastic attack problems.

Tim

unionised - 23-7-2006 at 21:08

Pure water is still a conductor anyway so I wouldn't try it.

IrC - 23-7-2006 at 22:03

Silicone oil is designed to be used at several hundred degrees without problems. This is why it is the choice in the fusor sections of high volume powder toner copying machines. Vegetable oil is lunacy and will rapidly start dissolving all the metal around it, meaning goodbye circuit traces and component leads. I put corn oil in a 1KW dummy load can and in just a few years it ate the steel container so paper thin you can push holes through the can with your finger that once would have needed a .22 cal CO2 pellot gun at very close range to puncture.

unionised - 24-7-2006 at 09:21

"Vegetable oil is lunacy " In the very real sense that it has worked under rather stressful condition (intermittent 20KV) for about 5 years.

Please don't call me a lunatic when I point out things that really are experimentally verified as true. Your oil may have rotted the metal; mine didn't. It's a good deal more interesting to check on what caused the difference than to call eachother names.

YT2095 - 24-7-2006 at 10:20

how about that oil used in Rod Welding transformers, that ought to work just fine too.

a Tool Hire shop may be able to point you in the right direction for getting some :)

IrC - 24-7-2006 at 11:26

""Vegetable oil is lunacy " In the very real sense that it has worked under rather stressful condition (intermittent 20KV) for about 5 years.

Please don't call me a lunatic when I point out things that really are experimentally verified as true. Your oil may have rotted the metal; mine didn't. It's a good deal more interesting to check on what caused the difference than to call eachother names."

Vegetable oils of the world unite! Your defender has arrived!

When your english comprehension skills improve perhaps you will read my post again and realize I was calling the idea lunacy and never mentioned any person by name nor did I call anyone a lunatic. What are you, the vegetable oil defender? Do you get upset when someone rags on a poor bottle of vegetable oil?

Must be so seeing as how you take my rag of vegetable oil as a personal name calling insult.

By the way give my your mailing address and I will send you a piece of the can, thin as tissue paper. Since this is a chemistry forum it never came to mind for me that anyone would not consider the heating and metal contact with vegetable oil over a long time as causing chemical reactions between the oil and the metal.

12AX7 - 24-7-2006 at 13:38

More to the point is, what reactions occured? Carboxylic, ester or -ene group complexation with metal ions? How would that lead to oxidation and dissolution of a solid metal? What kind of impurities were present in the oil which might lead to such a reaction?

Tim

IrC - 24-7-2006 at 14:24

I have no idea, didn't seem worth the investment to have a lab do a workup on my old oil. I can tell you I remember when the can was new, made from very tuff metal. You would have had to whang it a good one to dent the sides. I also remember when I noticed it leaking like a seive about 6 years later, and I also noticed that you could put a hole in the side with pressure from a finger. Taught me the value of Shell Diala-X. It was just a gallon bottle of pure corn oil from a Safeway store. I did not use the load that much, never long enough to even notice it ever really getting warm, meaning all this chemical activity took place at room temperatures over a period of around 6 years.

However, the oil is very dark and black and so is the can, both of which I still have if anyone is interested in doing a chemical workup on what was going on in there. When new, the can was mirror shiny and the oil was golden clear.

This much I know, it makes me wonder what is going on in my body when I eat anything deep fried in vegetable oil.

triphenylphosphineoxide - 28-7-2006 at 07:46

Welcome to the wonderfull world of free fatty acids.
Over heating vegetable oil produces them.
The vegetable oil itself is not a bad idea, assuming no contaminants, and protection from oxygen and heat. This thread made me check some olive oil I washed and dried last year. Thought I did the job properly, but the lid must not have been air tight as the oil is now bright green and the can is already suffering.
Personally I wouldn't want to risk this, happening to my PC.
Try Silicone oil, great thermal conductivity, won't degrade into anthing reactive, and shouldn't cause any solubuility problems within a PC's Life

unionised - 29-7-2006 at 08:06

Ho Hum,
Since lunacy is an abstract concept only applicable to sentient animals, the phrase "vegetable oil is lunacy" doesn't make any sense.
I presumed (and I think most people will have done) that you meant "the suggestion of using vegetable oil is lunacy". Since I made that suggestion (on the not altogether unerasonable gruonds that I know from experience, that it works) I took it as a criticism of me when you called it lunacy.
BTW, was the oil dried?

[Edited on 29-7-2006 by unionised]