Hello, I feel compelled to clarify some misleading comments here. I'm speaking as a former experimental physicist who routinely used liquid helium for cryogenic

Hello, I feel compelled to clarify some misleading comments here. I'm speaking as a former experimental physicist who routinely used liquid helium for cryogenic research. Firstly - BOTH isotopes of helium are important. Both are relatively rare, though He3 is much rarer. Both are needed for cryogenic research, though for MRI's He4 is generally sufficient (used for cooling the magnets, though a specialised lung imaging technique can use some He3). Both are byproducts of natural gas formation, and both will eventually escape Earth if released to atmosphere. (At earth surface temperasure these particles are so light that a significant fraction of the Maxwell Boltzmann velocity distribution is above escape velocity). Though I'm not sure what macroscopic time scale atmospheric helium will escape at. He4 evaporation temp is 4.2K so cooling down to that level is quite easy, just get a dewar of liquid He4 and dunk something into it. Realistically it's more complicated, you use carefully designed vacuum dewars with super insulation and usually a liquid nitrogen shroud (at 77K) to reduce heat transfer by radiation (which scales as T^4, meaning the LN2 shroud is roughly 1/4th room temp, and reduces radiative transfer by 250 times). You have carefully designed "dunker stick" with wires and frame carefully chosen to minimise heat transfer. Secondly, you can cool LHe4 a bit further to about 1K by evaporating the surface of the dewar. This evaporative cooling basically pumps away the more energetic gas particles in the upper tail of the velocity distribution, leaving the slower and cooler ones. This works but is very inefficient and would pump away most of your helium. To be more efficient you can use a "1K pot" to pump and cool only a small volume. Interesting things happen when you cool below the lambda point at 2.2K at which point the LHe4 becomes a superfluid. The pump noticeably gets quieter. Superfluous are cool but can also be a nuisance for cooling because they creep around and can form thermal bridges connecting parts of your system you wanted isolated and prevent further cooling. At this point you can get colder by using the MUCH rarer form of helium, He3 which has a lower boiling temp. You can evaporatively pump LHe3 and get down to about 200 mK. Since this isotope is so rare it's usually done in a tightly controlled closed fork system, the 'pump' can be a sealed can of activated charcoal IIRC. Things start getting interesting when you want to go colder. A mixture of both LHe3 and LHe4 becomes unstable in the right temp/pressure region and wants to form two distinct phases : a He3 rich and He3 dilute phase. A dilution refrigerator is a clever apparatus that uses hear two phases to run a closed-cycle refrigerator very much conceptually like your kitchen refrigerator. Evaporate He3 out of the rich phase, which cools, transfer heat away, compress back into the rich phase. A dilution refrigerator can get down to about 10 mK, maybe lower for ideal conditions. Though it's a beast. Can take all day to old and cool a sample, takes up half a room and just cools down a sample not much bigger than your thumb. Has several loud heavy vacuum pumps and huge dewar of LHe4, carefully controlled "mash" of He3/He4, etc. He3 can also form a superfluid at low enough temperature, and is interesting because in this case it's like Type I superconductivity where two fermions pair up and form a boson carrier. He4 atoms are bosons already but He3 are fermions.