# Cryogenic (LN2) gas vs liquid for cooling

• For cooling, would you recommend using gas over liquid or vice-versa? What is the temperature differential between gas and liquid?

In other words, at standard volumes and pressures from a gas discharge vs standard flows and pressure from the liquid dispenser side of the 180L tank,

• what kind of temperature differences can we expect?

I would expect the minimum achievable temp with a gas only system to be a lot higher than if we were feeding liquid directly to the setup, however I would also expect that using gas will help preserve the tank longer. Looking into Nitrogen phase diagrams now. I'm guessing that both gas and liquid are delivered at $22 PSI$.

• " What is the temperature differential between gas and liquid?" Are you asking for the latent heat? Also, please ask one question per post. Here you have at least two questions. The swagelok question isn't clear at all. Can you clean this up? – DanielSank Jan 6 '16 at 4:18
• Yeah once again, this is why I thought my original attempt with all questions and details was better, but everyone fussed at me to trim it down and separate all the questions into smaller groups. – Atomiklan Jan 6 '16 at 4:28
• Removed Swagelok question – Atomiklan Jan 6 '16 at 4:32
• I don't understand what you're saying about people fussing. I'm asking for you to make the question more focused, which is compatible with the previous request to "trim it down" as you say. – DanielSank Jan 6 '16 at 4:43
• That was from a previous massive post of lots of questions sorry. Dont worry about it anymore. – Atomiklan Jan 6 '16 at 4:44

At my workplaces (your mileage may vary), LN2 is so cheap and easily accessible, that it really doesn't matter. It's simply easier to dump a few liters of it wherever it's needed and not worry. In fact, that was the even general approach to handling LN2 back during my diploma studies in a lab where cryogens were generally in short supply.

However there are benefits to using a cryogen in gaseous form. The heat of vaporization of N2 is 5.56 kJ/mol. Its heat capacity is 29.1 J/(mol K). If you're cooling down something from room temperature, you can initially pull out some extra 29.1*(300-77)=6.5 kJ/mol of cooling power. Of course, that's no reason not to evaporate the cold LN2 against the object to be cooled, before using the vapors to cool it further.

One really starts to feel the difference when dealing with LHe though. It's beyond the scope of the OP, but the heat of vaporization of LHe is tiny. Tiny, as in, liquid helium cannot cool down anything. It serves to stabilize the low temperature of something, that's already cold and contained in a nicely insulated dewar/cryostat. It's the heat capacity of the gas that does the trick.

In that same lab I mentioned, we would very slowly fill a bath cryostat with LHe, making sure, that there was absolutely minimal overpressure of vaporized helium, so as to utilize its heat capacity as a gas to the fullest. Allow the pressure to rise, and the cold helium gas with all its cooling potential would escape into the recovery lines.

• Excellent answer, thanks. Sounds like our best bet is to just use the liquid withdraw. Yeah I think its about \$100 per tank swap through Praxair and it should take use quite awhile to burn through that tank. Should probably run some rough numbers to even see if we could use that amount in a typical month or if we're just going to lose it through tank bleed. – Atomiklan Jan 6 '16 at 5:56
• @Atomiklan It's hard to figure what your usage would be. As an example, I burned through 100 liters of LN2 in 10-14 consecutive days of work (top up an infrared bolometer twice daily, keep a little cold trap cold). Seems like most losses were indeed tank bleed and just carelessness. – LLlAMnYP Jan 6 '16 at 6:03
• Yeah I know these are kinda impossible questions without more data. Trying to do some basic flow calculations now to determine how quickly I would burn through a 180L tank using the liquid withdraw. 22PSIG to atmospheric through say a .25 Swagelok. Looks like it may actually go pretty quickly lol. – Atomiklan Jan 6 '16 at 6:07
• @Atomiklan see my comment in another question of your's then, something feels really wrong with using LN2 for continuous flow cooling. – LLlAMnYP Jan 6 '16 at 6:10
• Yes that was precisely my concern which was what I really wanted someone with experience to tell me. This gives us lots more to think about. Thank you. – Atomiklan Jan 6 '16 at 6:30