In September the US House of Representatives voted to delay closing the world's only strategic helium reserve, the Federal Helium Reserve. The world's materials scientists, who rely on liquid helium for a variety of research practices, breathed a collective sigh of relief. But the decision, whilst welcome, does not change the fact that helium is running out. If we are to continue low temperature materials research, we need to find alternatives.

To study electronic and magnetic properties and thus characterize new materials, researchers need to cool them to very low temperatures and subject them to high magnetic fields. Helium liquefies at 4 K, meaning liquid helium can be used to create the low temperatures required for many areas of materials study, as well as to cool the special coils in magnets to a superconducting state. A typical low temperature measurement system consists of a cooling platform (traditionally using liquid helium), a space to place the materials sample where temperature can be varied, and a set of probes to measure electrical, magnetic and thermal properties. Similar conditions are also often required for materials being studied on beamlines at neutron or light sources.

Graphene is perhaps the most high profile example of a new, exciting material. Researchers, including Konstantin Novoselov and Andre Geim at the University of Manchester who won the Nobel Prize for their graphene work, regularly use these types of systems to characterize many aspects of this supermaterial.

Cooling the samples often requires reaching close to absolute zero. The traditional way of cooling has been to continuously pump liquid helium around the system to draw heat away from the cooled area. Special versions of these refrigerators can reach temperature of only millikelvin from absolute zero. All however start by using liquid helium at 4 K.

Helium is produced as a by-product of extracting natural gas, and is used in a number of industrial applications such as mixing with oxygen for deep sea divers as well as in scientific instruments and superconducting magnets. Once helium is released into the air it is lost forever, and few natural gas wells are in a position to produce more helium in an economically viable way. So we are reliant on our limited existing supplies.

The Federal Helium Reserve in Amarillo, which provides 42% of the country's helium and 30% of the world's, is one major source. The fact that it is remaining open is good news for the community. But it is a reserve that is not being replenished and will run out – probably within 25 years. And as supply dwindles, the price will rise rapidly. In addition to decreasing supplies and rising prices, liquefaction of helium also requires a lot of energy. In the UK we pay around £6 to purchase a litre of liquid helium, but in Japan – a world leader in low temperature research – this rises to over £20. Regions with emerging research bases such as the Middle East, Nigeria and Brazil, fare even worse. They do not have the facilities to produce liquid helium and must import it, an even costlier operation – especially if it sits in customs for weeks boiling away.

Superconducting magnets which use liquid helium also come with their own challenges. Regularly topping up machines is a complicated process, and magnets can occasionally quench – a process in which the liquid helium surrounding the magnet rapidly boils off. Systems using liquid helium require large areas with suitable ventilation facilities and a technician on hand who is trained in cryogenics.

So, whilst the days of liquid helium are not quite over, it is clearly becoming less and less attractive.

One option has been to try to recover the gas as it boils off, but this can be costly and, except in the most advanced systems such as the LHC at CERN, users struggle to capture, purify and then reliquify 100% of the gas. Having spent a lifetime in cryogenics, I’m convinced the future lies in ‘cryogen-free’ or ‘dry’ mechanical systems.

Unlike existing systems, cryogen free technology uses mechanical refrigerators consisting of a compressor and cold head package. These cool to cryogenic temperatures using only electrical power. They are heat engines and use Gifford-McMahon (GM) or a Pulse Tube (PT) thermodynamic cycle to provide cooling. The cycle involves repeated compression and expansion of a small quantity of helium gas to generate low temperatures.

They have slightly different thermodynamic cycles but in both cases the gas is supplied by an external conventional compressor, a motor driven set of valves causes repeated expansion of the gas causing a decrease in both pressure and temperature. The GM machine has moving cold pistons through which the gas passes. The PT machine relies on a gas pulse resonance to control and set the flow through regenerating heat exchangers. It has no cold moving parts which many people consider an advantage. Finally, the gas returns to the compressor, completing the closed loop circulation of gas through the cold head. The level of helium required is extremely small and no gas is lost in the process.

Both GM and PT coolers can reach temperatures as low as 2.6 K, and have a cooling power of 1–2 W at 4 K. They can run for more than a year without attention, making the use of low temperatures very simple for technicians without specialist knowledge. The upshot is that low temperatures can be created anywhere with just an electrical power supply.

Purchasing large amounts of liquid helium is becoming less and less viable. In recent years major research laboratories have had to temporarily shut down multimillion-pound facilities because of these shortages and the problem will only get worse.

Providing an alternative which does not rely on a regular supply of helium or an expensive liquefaction process, is essential if we are to continue and expand global materials research. Not only will it address the rising cost and depleting supplies of helium, but it will make low temperature research viable in emerging research areas which cannot access, or cannot afford, liquid helium.

Read full text on ScienceDirect

DOI: 10.1016/j.mattod.2013.12.001