WE GET to grips with temperature at an early age. Parental rites ensure a baby’s room is stiflingly warm, its bathwater “just right”, and that some things are “Hot! Don’t touch”.
As we grow older, we begin to associate numbers with temperature sensations. We learn that at 0 °C water freezes, that 20 °C describes a warm day and that 37 °C is a biochemical Mecca. At some point in our growing awareness, perhaps at school, we may even come across a temperature far beyond our everyday ken: absolute zero.
Absolute zero is the ultimate in cool, an ideal, unattainably perfect state of coldness. Even so, since the concept first emerged in the mid-19th century, many careers have been devoted to getting ever closer to it. What might seem a quixotic quest – noble, but pointless – is in fact far from it. This year marks the 100th anniversary of the first of many physics Nobel prizes awarded on the back of it. Each advance towards absolute zero has uncovered new states of unparalleled beauty and order, led to engineering marvels and enhanced our insights into basic science – not least about notions of temperature and matter itself.
Advertisement
Temperature’s familiarity makes it easy to miss how puzzling the concept can appear. Early natural philosophers such as Galileo, Isaac Newton and Robert Boyle thought heat a kind of fluid, called caloric, and we still speak of heat “flowing”. Others thought that cold was caused by the presence of “frigorific atoms”.
Early attempts to reliably measure heat and temperature were similarly fraught. The most useful early thermometers exploited the way liquids expand as they get hotter. A liquid was constrained in a glass bulb or narrow tube, and its level was marked at two fixed points: when exposed to boiling water and to melting ice, for example. Unknown temperatures were measured as “degrees of heat” etched as a scale between the two. This led to a catch-22: the calibration process presumed that the liquid expanded an equal amount for every unit rise in temperature, but this assumption couldn’t be verified without measuring the expansion of the liquid as temperature rose, for which one required… a thermometer.
Only exhaustive experiments by the French scientist Henri Victor Regnault in the 1840s, using a thermometer that measured changes in the pressure of dry air in a sealed container, established reliably reproducible temperature readings. This was a boon for science and industry – but it still didn’t really explain what was being measured.
The confusion was evident in the plethora of early scales used to measure temperature. Some of these, such as the Celsius and Fahrenheit scales calibrated using various properties of water, are still with us today. It was the 19th century physicist William Thomson, later to become Lord Kelvin, who first considered the possibility of an absolute temperature scale that didn’t depend on the properties of any one material. His own recipe was obscure, resting on the operation of the ideal heat engine first imagined by the French scientist Nicolas Léonard Sadi Carnot. But a more powerful – and ultimately successful – explanatory concept was emerging.
It is hard to imagine a time when the great scientific pioneers didn’t understand that everyday objects are made of atoms. With that understanding, the nature of temperature can become clear. Heat is the kinetic energy of motion of atoms, and temperature a measure of the speed with which atoms move – more exactly, the square of the average molecular speed. When we experience the temperature of a substance in our everyday lives, we are literally sensing the “buzzing” of matter.
Once the idea of molecules jiggling within a substance is accepted, it gives the absolute zero of temperature a natural definition: it is the point at which atoms become completely still. The question then becomes, “at what temperature does that happen?”
Clues were around for those in the know. Guillaume Amontons, a 17th century French instrument-maker, noted how the pressure of gas sealed in a vessel fell by “around a quarter” when the gas was cooled from water’s boiling point to its freezing point. Extrapolating, he speculated that if the gas were cooled further, its pressure might eventually disappear altogether at a temperature we would now describe as -300 °C. That is not far off, as more accurate investigations of the pressure and temperature of ideal gases have since shown. Absolute zero – now known as the zero of the Kelvin temperature scale – lies at -273.15 °C.
A never-ending story
The race to get there began in earnest in the late 19th century. Like the race to Earth’s frigid poles that was happening around the same time, it was a journey into the unknown. But the goal of this race could never be reached.
“Like the race to Earth’s poles, the search for absolute zero was a journey into the unknown”
One way to understand why is to consider how a conventional refrigerator works. Its interior is placed in contact with a colder substance, typically a circulating fluid, so that heat flows out into the fluid, cooling the fridge’s contents. To get heat to flow from an object that you want to reach absolute zero, the fluid coolant would have to be colder than 0 K. That is a nonsense: it is impossible to make molecules move slower than not moving at all. The best you can hope for is to still them as much as possible.
The coolant in a refrigerator is cooled by expanding it, lowering its pressure and thus reducing the average speed at which its molecules move. This same technique was used in the first stages of the race to the bottom. One gas after another was cooled under pressure, before being allowed to expand rapidly. This lowered its temperature further and caused it to condense, changing from gas to liquid.
In the late 1870s, the Frenchman Louis-Paul Cailletet used cascades of expanding gases to liquefy oxygen at -183 °C and nitrogen at -196 °C. It is doubtful he or anyone else realised how commonplace these substances would become in the 20th century. If applications were envisaged, I would bet that destroying warts and making instant ice cream wouldn’t have been among them.
After the liquefaction of hydrogen, by Scotsman James Dewar in 1898 at -250 °C, there remained only helium, whose weakly interacting atoms make it the hardest gas to condense. The ingenuity and effort required to persuade helium atoms to become a liquid were eventually amply rewarded. On 10 July 1908, Heike Kamerlingh Onnes at the University of Leiden in the Netherlands reached a temperature of 4.2 K and produced the first few cubic centimetres of liquid helium.
A whole new world
The liquid opened the door to a whole new world of physics. Shortly after liquefying helium, Kamerlingh Onnes discovered that at very low temperatures some metals become superconductors. On cooling below a critical temperature their electrical resistivity falls by at least 15 orders of magnitude to a value indistinguishable from zero. It wasn’t long before the Nobel prize committee recognised the value of Kamerlingh Onnes’s work, awarding him the . Even if superconducting technology is not yet as commonplace as many had hoped it would become, it is used in the magnets of MRI body scanners, and for the hugely powerful magnets that bend beams of protons round the circumference of the Large Hadron Collider near Geneva, Switzerland.
Perhaps the most astonishing low-temperature phenomenon, however, was one that took place in front of Kamerlingh Onnes’s eyes on that first day of liquefaction. Peering through small gaps in the insulated glass vessel, he could see the almost perfectly transparent liquid boiling. By sucking out helium vapour from the space above the liquid, the fastest helium molecules could be removed and the liquid cooled even further, yet the vigour of the boiling increased.
But then, below a temperature we now identify as 2.17 K, the bubbling suddenly stopped and the liquid became eerily still. It took years before we discovered what was going on. A fraction of the liquid had changed to a new state, a superfluid, that flows without resistance and conducts heat perfectly. Whenever a region of the liquid became marginally hotter and began to form a bubble, the superfluid carried the heat away before the bubble could form.
Other astonishing insights followed. Helium nuclei usually contain two neutrons as well as two protons, and therefore the most common form of the atom is named helium-4. Thousands of times rarer is the isotope helium-3, which has only a single neutron. These lighter atoms condensed at 3.2 K, rather than 4.2 K, and once liquefied behaved completely differently, becoming more rather than less viscous as they cooled, for example.
Who would have guessed that the presence or absence of a single neutron could so transform the physical properties of a liquid? I would call such phenomena extraordinary, but they are not – they are completely ordinary. We are just unaware of how astonishing ordinary matter is.
The common truth behind these seemingly bizarre behaviours is that the world in which we live is described by quantum mechanics. Only when random thermal vibrations are reduced by cooling does this reality become apparent. We see, for example, that the interactions between helium atoms become so weak that quantum mechanics allows them to swap places without the inconvenience of going around each other. This quantum swapping causes both types of helium to remain liquid to the lowest temperatures investigated. In fact calculations indicate that at atmospheric pressure, helium would remain a liquid even at absolute zero.
Investigating quantum phenomena is one reason for pushing further towards absolute zero. The differing properties of helium-3 and helium-4 when liquefied provide us with the next stepping stone. In a device known as a dilution refrigerator, the superfluidity of helium-4 causes liquid helium-3 to behave like a gas, effectively evaporating into a helium-4 “vacuum” and cooling the set-up as far down as 0.001 K, or 1 millikelvin. At these temperatures helium-3 itself finally becomes a superfluid – but a magnetic one.
If cooling to 0.001 K is hard, going further is incredibly difficult. The thermal conductivity of all materials plummets as the temperature falls, meaning it takes ever longer to remove heat from a substance. At the same time the heat capacity of all materials – the amount of heat needed to change its temperature – becomes tiny, so any experimental technique used to study a substance will warm it up. If something as delicate as a butterfly were to fall 10 centimetres onto a cubic centimetre of copper at 0.001 K, the impact’s energy would raise the copper’s temperature 100-fold.
There are workarounds, at least for smaller amounts of material. One that works for a gas of up to a billion or so atoms is to cool atoms individually. Laser photons collide with each atom, carrying away momentum and damping the atom’s motion. The principle is still to carry away one body’s heat using another; it is just the sophistication of the coolant that has changed. With this method, we can slow atoms from moving at around 1 metre per second at 1 millikelvin to roughly 1 millimetre per second at 1 nanokelvin.
The reward for such ingenuity is the chance to explore how matter behaves when it is dominated by quantum mechanics rather than chaotic thermal disorder. We know that superconductivity is fundamentally a quantum phenomenon, for example, yet despite investing billions of dollars we still don’t know how some superconductors work at temperatures as high as 130 K. By creating a more controllable quantum system we can use a gas of super-cooled atoms as a simulator to investigate such phenomena, probing and stimulating the atoms’ interactions using pulses of laser light.
We can also use the incomparably pure quantum-dominated environments of ultra-cold materials to model extreme conditions in the interior of a neutron star, fundamental particle interactions, and phase transitions in the earliest moments of the universe. At low temperatures, electrons interact to create fundamental excitations – sometimes called quasiparticles – with a mass up to a thousand times that of a free electron, much as fundamental particles in free space acquire mass by interaction with the Higgs field. Similarly, quasiparticle excitations in superconductors have recently been shown to behave like Majorana particles, long-predicted objects that are their own antiparticles.
Direct applications resulting from these experiments seem far-off at the moment. Given the last century of progress, however, we would be unwise to bet against them. Majorana fermions are one of many particles that have been proposed as possible processing bits for future quantum computers. Realising the potential of such machines, whose data-crunching abilities would put conventional computers in the shade, may depend on our mastering the intricacies of matter at low temperature.
It might seem odd that, a century after Kamerlingh Onnes took us to 4.2 K, we are still investigating what happens in that last degree above absolute zero. We shouldn’t think about that single degree, however, but instead about the factor 1000 difference in temperature between 1 K and 1 millikelvin, 1 millikelvin and 1 microkelvin, 1 microkelvin and 1 nanokelvin.
Each stage of cooling is like lowering the background noise in a room, allowing us to hear ever quieter natural voices. Cooling further allows us to probe atomic interactions at new levels of subtlety. Even at 1 nanokelvin, there is plenty of room at the bottom – to picokelvin, femtokelvin and beyond. The story so far tells us we really have no idea what we will find when we get there.
“Each stage of cooling is like lowering the noise in a room, allowing us to hear ever quieter natural voices”
This article appeared in print under the headline “Absolutely freezing”
