Water vs. Ammonia (for Sulfur)

[Naeleros]

Sulfur..

I couldn't find anything in that mess of a thread. So.. I made a new one.

Anyway, water cools better than ammonia. However, water has a much higher freezing point than ammonia. So.... if you're going into very low temps then you need to consider the possibility of your liquid freezing. That is why some people use ammonia (usually a mixture of ammonia and water) in their coolers.

Hope that helps.

James

[Crazy Jayhawk]

Another advantage for ammonia would be that it could work as an antibacterial agent. Most household cleaning chemicals use ammonia after all.

[[GeHlGe]R@g!ng]

how to get the water(amonia) below 0 c?
i'm looking for some sollutions to to chill my cpu.

[SUlfUR-CyCLoNe]

Thank god somebody looked into this for me!!!!

What I'm trying to do is use a dual pelt system that is water cooled. I'm also going to out in two 92's in the front and 2 80's in the back for incresed case cooling.... I might end up only using one peltier since two is pretty tricky.

-SUlfUR

[Karsta]

As you probably know ammonia is gaseous at normal temperatures and pressure. It has boiling point of -33.5°C (-28.3°F). It is easily liquefied with pressure which is the reason it has been used as a refridgerant.

Anyway it is very nasty stuff: Highly toxic and irritant (and has very strong stink). It has moderate fire risk but it may form explosive compounds in contact with silver or mercury.
Consentrated ammonia solution (ammonium hydroxide) has also very strong odor and is extremely irritating (esp. eyes).

If the reason for using ammonium hydroxide would be to get sub zero temperatures, I would advice using something user friendlier stuff like ethanol, glycol etc. (which also hinder unwanted life in tubing). For refridgerators there has been safer gases (doesn't mean safer to ozone layer though) than anhydrous ammonia available for long time.

[Naeleros]

I believe that anhydrous ammonia is a specific type of ammonia and not the type that they use in watercoolers.

I am definitely no chemistry expert... so.. I reserve the right to be totally wrong. But... anhydrous ammonia is a chemical that is used in farm fertilization (among other things) here in the US. I would HIGHLY recommend that you not handle it unless you have a good idea of what you are doing.

Anhydrous is gaseous in its natural state and converts to liquid under pressure (as you already stated). It expands from liquid to gas at 800 to 1... so.. a small spill is a huge gas cloud. It can cause serious damage to human respiration process. It can also burn the skin. On top of all that.. it is very cold in liquid form. Cold enough to injure.. I believe.

I had assumed.. that there was a different type of ammonia, though. Surely... surely... no one is using anhydrous for watercooling????

James

[Karsta]

I mentioned also water solution of ammonia (ammonium hydroxide). Ammonia evaporates from water so if it is used the system must be closed and filling of system and servicing etc. has to be done in a well ventilated place (fume hood if inside).

Other thing is that ammonium hydroxide is basic. Aluminium reacts with it producing hydrogen gas and aluminium hydroxide salts. So one could not use aluminium blocks or tubing. Ammonium hydroxide solution also conducts electricity which is bad. If one uses dilute ammonium hydroxide, it is not as dangerous for health but nevertheless, not well suited for water cooling.

Liquid ammonia doesn't have to be cold. It has to be kept in pressure vessel or tubing to hinder evaporation. For example I have done reactions in liquid ammonia with temperature of about 75°C. Liquid ammonia has been used for cooling as I mentioned. I have never heard that water solution of ammonia would have any advantages though.

Using ammonia or ammonium hydroxide doesn't offer but risks and harm in home environment, so I wouldn't recommend using them.

[r0ckstarbob]

hm

what thread was this initially a part of?

anyhow

in the search for the perfect sub-Z coolant, this is the gunk that i kinda learned

it seems the simpler the molecule, the better the HT properties. i haven't found any documentation to support this but it seems the best chemicals for Heat Transfer tend to be very simple molecules. i'm sure if i dug around a little more, i could probably find something to that effect. i dunno.

anyhow...

Water -
makeup = H20
freezing point = 32F

has some of the best heat-transfer properties on earth and is just about the best thing to use (beyond using liquid metal like the russians do for their reactors). the only thing that improves water is waterwetter or purple ice which will increase the HT properties of water by up to 15-20% with the proper ratio by breaking up the surface tention. whatever coolant solution you decide on, it should be used to cut water. remember, the more we cut the water with another compound, the less efficient it's going to be. so we're looking to have a mixture that uses the most water to "other" ratio to find the ideal compound.

Glycol / Antifreeze - it's a silicone.
probably the worst gunk you could put in your system as it has virtually no heat transfer properties (is almost liquid styrofoam). it's good for raising the boiling point of water and is okay at lowering the freezing point but will still slush up at around -10F. it's main bennefits are that it prevents wear and tear on yer stuff by coating it in a thin layer of glycol (silicone) and it prevents bacterial growth (stuff can't grow in a petrolium compound). it works by insulating H20 molecules. it's like putting a down-feather coat around each molecule. it also has terrible viscocity which means slower flow rates, harder work for your pumps to push.

Alcohol
theres lots of different types of Alcohol

Isopropyl (rubbing alcohol)-
makeup = CH3CHCH3
freezing point = -127F
is a pretty complex molecule and is very "processed". it's this complexity that kinda steered me away from it.

Methyl/Menthanol (wood alcohol) -
makeup = CH3OH
freezing point = -144F
the simplest of all the alcohol molecules. it's more processed then ethanol but the result is to simplify the molecule.

Ethyl/Ethanol (grain alcohol) -
makup = CH3CH2OH
freezing point = -173.4F
used to make drinking alcohol. this is what you get when you distill alcohol using traditional methods and is considered "pure" alcohol. they denature it (make it undrinkable) by cutting it with a small amount of methanol

the cool thing about alcohol is that it has no detectable PH balance which means it's non-corrosive. it's also got a VERY high viscocity, which means much better flow. for myself, i've decided to run with a methanol/water mixture due to the fact of the simplicity of the molecule. with an increased viscocity (like 3 TIMES as viscous as water which translates to loosely 3 times as much liquid) it is possible to make up for whatever loss in heat transfer properties may occur by moving more liquid over it. it's a quantity vs quality issue. it is possible that ethanol may be a better choice. the thought being that if the freezing point of ethanol is lower then methanol, that means it's more potent, which would in turn mean that you would need less of it (to water) attain the same effects. remember, the more water the better. again, i haven't researched this theory but think it may have some foundation in fact.

i haven't explored Ammonia much except used with a refrigerant, due in large part to it's PH balance which i think is quite high (though i don't know) and worried about it's corrosive properties. also looked into chlorine and bleach as a possible coolant. have seriously considered using Butane but i get all jittery using a highly combustable fuel, however good the temps.

something else i looked into and liked...

hydrogen peroxide.
makeup = H202
freezing point = -70F
industrial strength (50:50 water) has much better HT properties then a water/alcohol mix and doesn't react at all with aluminum, but untreated, will eat through copper in a heart beat. you can treat it with Nitric Acid, but industrial strength (treated) Hydrogen Peroxide is an item that i'm forseeing difficulty in procuring. if you ran an entirely aluminum system, i think this would be an ideal choice, but again, something i've never heard of anyone else doing and haven't had the resources to check into myself.

i dunno. stuff to think about and good stuff to know i think. especially when ventuing into the subZ category of this mess.

peace

RSB

:cool:

[Karsta]

I appreciate your efforts on water cooling. I have not gone to water cooling yet so I lack all the practical knowledge. There is some things I feel I have to correct though:

Glycol / Antifreeze - it's a silicone.
probably the worst gunk you could put in your system as it has virtually no heat transfer properties (is almost liquid styrofoam).

Glycol is 1,2-ethanediol (HOCH2CH2OH). It has nothing to do with silicone. Word glycol is also used as a common name for dihydric alcohols. It is very viscose and because of that one needs higher pressures for good flow. Anyway it is commonly used coolant and antifreeze. I don't know about its heat transfer properties but because it is widely used they can't be so poor.

it's also got a VERY high viscocity, which means much better flow. for myself, i've decided to run with a methanol/water mixture due to the fact of the simplicity of the molecule. with an increased viscocity (like 3 TIMES as viscous as water which translates to loosely 3 times as much liquid) it is possible to make up for whatever loss in heat transfer properties may occur by moving more liquid over it.

On contrary high viscosity means lower flow rate or higher pressure needed. With glycol you thought the viscosity right way. Here are some viscosity coefficients of pure compounds:
methanol 5.97x10^-4 Nsm^-2,
water 8.90x10^-4 Nsm^-2,
ethanol 11.90x10^-4 Nsm^-2,
glycol 173x10^-4 Nsm^-2

I would be very careful with hydrogen peroxide. Concentrated solutions are highly toxic and strong irritants. Hydrogen peroxide is a unstable compound and a strong oxidizing agent. It would also bring a serious fire and explosion risk to your computer...

And yet about butane which is gas. It is a simple hydrocarbon still well soluble in water. There is of course a fire risk if it leaks. I would think that it is used as a refridgerant in pure form (like ammonia, CFC and HCFC compounds) and not so much as water solution.

Use of chlorine doesn't sound a healthy idea to me.

One note: commonly all impurities in a solvent lower its freezing point. So it is also possible to use also solid materials as additives (as far as they are soluble and won't gradually crystallize in the tubing).

[venFiken]

Vodka! Yeah I know is sounds corny but think about it. Go stick a 1/5'th of Mr.B in the freezer and let it sit for a couple of years. Now pour it out and notice no ice whatsoever.

Does it work? I dont know about sub zero solutions but friend tim has has smirnoff running in his coolance <-guess for a couple months with no effects.....

[r0ckstarbob]

wow. okay. very cool. maybe put together with my knowlege of cooling and your chemistry knowlege we can come up with something. my chemistry knowlege is at an all time low and it wasn't until i started really looking into subZ coolants that i had to learn. seems that is the way with alot of things no?

Originally posted by Karsta
Glycol is 1,2-ethanediol (HOCH2CH2OH). It has nothing to do with silicone. Word glycol is also used as a common name for dihydric alcohols. It is very viscose and because of that one needs higher pressures for good flow. Anyway it is commonly used coolant and antifreeze. I don't know about its heat transfer properties but because it is widely used they can't be so poor.

hm. this appears to be my bad. i dunno where i got that idea that glycol is a silicone and now i've lost my bookmarks somehow. maybe i never had them. hard to say. but you're right. ethylene glycol and propylene glycol are not silicones afterall. i do know both their heat transfer properties are absolutely terrible, that it works by insulating the water against temperature change (raises the boiling point and lowers the freezing point). cars like it for that reason. not because it conducts heat well, but because it can continue to use water above the boiling point. the term antifreeze is misleading. it should be called antiboil because that is where the emphasis is when they design anti-freeze. while it does lower the freezing point, it seems to be almost an added benefit, and something they hadn't counted on. its good for cars using and in automotive applications where it won't have to stay cold for long (once you start the car it will warm up and begin to be effective- in a car, the coolant only has to not turn into a sold so the pump can turn it over)

http://www.agaweb.com/coolcpu/faq2.htm#3
http://www.redlineoil.com/redlineoil/wwti.htm

On contrary high viscosity means lower flow rate or higher pressure needed. With glycol you thought the viscosity right way. Here are some viscosity coefficients of pure compounds:
methanol 5.97x10^-4 Nsm^-2,
water 8.90x10^-4 Nsm^-2,
ethanol 11.90x10^-4 Nsm^-2,
glycol 173x10^-4 Nsm^-2

this appears to be my bad again on the terminology. lower viscocity is definately better. knew that, just used the wrong phrase. didn't know the mathematical viscocity coefficients exactly but have been looking for them for weeks. it's likely i didn't know what i was looking at. by looking at your figures though, it would seem that my theory of maybe using ethanol just got debunked. if i understand your figures correctly, ethanol has a higher viscocity then even water, where methanol has a lower viscocity and thus would work better. i think it may be likely that the higher viscocity of ethanol would probably contribute to it's lower freezing point. i dunno. am just speculating. as we can see though, glycol has almost 20 TIMES the viscocity of water... that is unless you forgot to include a decimal point and that was suppose to be 17.3... at any rate, thats still almost twice as viscous... reason enough to leave it alone and look for other solutions me thinks. methanol has better HT properties then glycol and better viscocity. cut with water, it seems to be the ideal solution.

i got the figure of 3 times from this website. didn't actually do the math myself.

http://bioen.okstate.edu/home/mjwhite/cooling/project1/
and this is just a good link to have around
http://www.agaweb.com/coolcpu/science.htm

I would be very careful with hydrogen peroxide. Concentrated solutions are highly toxic and strong irritants. Hydrogen peroxide is a unstable compound and a strong oxidizing agent. It would also bring a serious fire and explosion risk to your computer...

And yet about butane which is gas. It is a simple hydrocarbon still well soluble in water. There is of course a fire risk if it leaks. I would think that it is used as a refridgerant in pure form (like ammonia, CFC and HCFC compounds) and not so much as water solution.

Use of chlorine doesn't sound a healthy idea to me.

One note: commonly all impurities in a solvent lower its freezing point. So it is also possible to use also solid materials as additives (as far as they are soluble and won't gradually crystallize in the tubing). [/B]

you're right of course using things like hydrogen peroxide and chlorine... especially in reference to having to use industrial strength 50:50 mix. if i get REAL serious, i may start looking into it down the road, but for now, i was just exploring the option.

the idea of using solid materials in your liquid is correct but i don't think the most efficient idea. besides the fact that your pump will have to work overtime to pump it with the liquid, if we acknowlege that water is the best heat transfer stuff around, then every piece of solid material in your water isn't transferring heat, or isn't transferring it as well as water. the idea i think is to find a compound that lowers the freezing point and still effects some amount of heat transfer in the process to make up for the fact that you're using a less efficient heat transfer compound then water. at a 50:50 ratio of water to glycol the heat transfer properties are halved. using a 50:50 ratio of methanol/water, the heat transfer properties are only reduced by 15-20% if i recall. (will try to find the figures where i got that number, they may be off, you may want to check). combined with a better viscocity leading to higher flow rates and a 3% waterwetter (as recommended to me by the waterwetter company when i emailed them about this) i'm thinkin we might have a winner.

i'm glad you brought the idea of using solids in the liquid to effect the change though. maybe someone has an idea along those lines that works better still. who knows. i'm glad someone brought it up regardless.

thanks for the info. it really is incredibly helpful.

RSB

[Karsta]

Sorry for late response. I have been (and am a bit busy).

I am glad to help if can. My chemistry knowledge is thin and general what comes to cooling systems. I am a chemist but working with peptide chemistry and such. Anyway I feel interest towards water cooling or some other cooling system with better efficiency/noise ratio than air cooling. When I start seriously to get this thing overclocked I probably will have to change cooling. Copper blocks and other necessary stuff is quite expensive to buy so I will maybe try to make as much as I can things myself.

To me methanol sounds a very good additive for coolant if used in closed system. I right now think that it will be my choice also - if somebody doesn't come up with something even better.

To solid materials - I meant stuff that is fully soluble in coolant (water). When molecule is solvated it doesn't act anymore as solid. Also the properties of water will change (density, freezing and boiling point, conductivity etc.). The positive effect for such additives could be antimicrobial (good for open system, because there wouldn't be evaporating poisonous material) and lowering of the freezing point. About heat conductivity I don't know. Downside with salts is of course conductivity, and possible effect to pH. So the additive should be some neutral molecule.

I stumbled to an interesting paper. Writers used solid nanoparticles made from Al2O3 as an additive for coolant. They got 20-30% positive effect on cooling efficiency. Article was written last year and nanofluid is far from commercial product though (can't do them home).

[r0ckstarbob]

hmmm about solids

completely soluable huh? like you, first thought was salt. but again, you mirrored my concerns.

so what else? sugar?

how well does sugar water transfer heat?

could there be something to using traditional alcohol with their impurities and their sugars that actually increase it's HT properties as a coolant?

should our systems be cooled with Black Velvet or Smirnoff?

jeez. i mean, i guess it's possible right? anyone know for certain?

i don't think it's electrical conductivity should be a worry unless it has sme direct bearing on it's heat transfer properties. does it? i don't know for certain. we're using a closed system anyhow. all should be fine me thinks...

here, check out this link. more talk about stuff like this here on the fourms.

http://forums.overclockers.ws/vb/showthread.php?s=&postid=268617&t=1488#post268617




thinkin thinkin thinkin....

[Karsta]

Wow! I Found link to your project. You are in to an very impressive cooling system. You have thought it very thoroughly. What kind of refridgeration you are going to use? Have you started to build it yet?

I think that using only solid additives, it would be very hard to get so low temperatures with liquid water. (with certain salts and ice you can go even lower. Solvation uses the energy - not a continuous system) If you put lots of sugar or like in the coolant it would probably crystallize in the plumbing. With some compounds it may be also possible that water and ethanol will separate to two phases which would of course be a very bad thing. I have once seen something like this happening for some liqueur in freezer.

Your water cooling system will be very efficient because surrounding air will keep near the coolant temperature at the cold end. Because of that the heat conductivity of the coolant is not so critical. To ensure that air moves enough inside the case you could add a noiseless fan which would be on all the time computer is on.

Because of low temperatures and methanol/ethanol there will not be any microbial growth in the plumbing.

I think there was a thread about consequences electrical conductivity of water some time ago. I don't remember where the discussion lead though.

[r0ckstarbob]

hey wow. thanks! i'm glad you like it! no, its still in the design stage and has been for about 4 months now. i'm almost done now organizing my notes and sketches into something readable and will be cutting a set of detailed diagrams and building plans/specs within the next two weeks or so. would really like to use Ammonia as the refrigerant but might end up using Freon if we have to. luckily i'm building it by hand so will be able to build it to spec one way or the other and hope to be able to make up for whatever loss may occur in having to go over to freon with a combination of specialty design mods and maybe a more powerful compressor. we'll see.

have considered just keeping the scrubber running on the inside the whole time to keep air circulating around. don't see why it would be a problem. it wouldn't be depeleting the drierite any more then normal. once it's sucked up all the water it can (will run out of humid air before i run out of drierite easily) it's not going to be noisy beyond the sound of the compressor going and perhaps a soft humming from the scrubber fans running on the inside... i mean its going to be an air-tight enclosure and well insulated so...

but you're right, good air circulation insde the core is going to be pretty critical, and luckily, pretty easy to establish. will definately do something along those lines. thermal conductivity of the liquid also won't be as crucial due to the design of it i know, but it's likely to play a factor down the road after we go ahead and produce this first prototype and then later decide to refine it in model number 2. if i can tackle it now, i'm going to try. :)

alright, back to brass tacks

okay this just gets weirder and weirder but hopefully some good will come of it.

what about something else as a desolving solid? Niacin or maybe Antacid like Alka Seltzer? baking soda?

just curiously, on an incredibly strange note, i wonder if enough common Antacid would negate the corrosive effects of hydrogen peroxide. i know it's a wierd thought and way out there in left field but sometimes...

anyhow, apparently theres a way to dumb it down and kill the corrosive properties of it so you can use it. apparently they use a nitric acid to treat H202 so it wont react against things like copper. it's got great characteristics from what i can tell.

oh, and while i'm thinking about it. am looking for the viscocity (on the same scale as you provided me) for the following chemicals. if you know it off the top of your head, that would rock. failing that i'll find the information. this is merely for the website and my notes.

looking for the viscocity of:
air
water/glycol (50:50) ratio
isopropyl
hydrogen peroxide

and looking for the heat conductivity (watt/m - per Centigrade) and heat capacity (cal/g - per Centigrade) for

isopropyl
methanol
ehtanol
h202
gold
silver

am building a table and hopefully be able to compile this in my notes and eventually on the webpage. maybe it'll do someone some good one day. who knows? thats the hope.

thanks again. this is fun.

RSB

[joshiieeii]

Originally posted by r0ckstarbob
hmmm about solids

completely soluable huh? like you, first thought was salt. but again, you mirrored my concerns.

so what else? sugar?

how well does sugar water transfer heat?
Ummm... No, heh, unless you want your puter to manufacture sugar crystal farms (how you think "After Shock" got those lil' crystals in it??)

Originally posted by r0ckstarbob
could there be something to using traditional alcohol with their impurities and their sugars that actually increase it's HT properties as a coolant?

should our systems be cooled with Black Velvet or Smirnoff?

jeez. i mean, i guess it's possible right? anyone know for certain?.....

You should use DIESEL, its 200 proof!! (thats 100% alcohol :D )

[Overclockers]

u can fill your reservoir with urine:D it will work great and help the recycling program as well
because urine contain water and amonia.:D

[phoenixpinion]

wouldnt it be much easier to have a second radiator in a freezer right before the water goes to your cpu

[r0ckstarbob]

*chuckles*

[r0ckstarbob]

alrighty boys

just made a new thread with a bunch of technical data on the main cooling forum. Sulfur & Naeleros, it's probably going to be right up your alley.



CHECK IT OUT (http://forums.overclockers.ws/vb/showthread.php?s=&threadid=39039)