TO IMAGINE working your way down the temperature scale into the realms of extreme cold, you might start with the inside of an industrial freezer. At about -18°C (0°F), it is uncomfortable, but bearable with some warm clothes. Now, turn your mind to the -60°C (-76°F) or so that explorers experience during the Antarctic winter, a temperature so severe it can ruin human flesh. And then, for the ultimate in cold, think of space itself at -270°C.
Except, it might surprise you to hear that the coldest places in the universe aren’t in space, but in the physics departments of many universities. Here, over the past few decades, researchers have been contriving ways to reach ever closer to the coldest possible temperature, absolute zero. In the process, we have entered a new realm where bizarre states of matter can flow uphill, chemical reactions can be paused and designer materials can be assembled.
Now, though, the conquest of cold is entering a new phase as we build the ultimate cooling technology: the quantum fridge. Power up one of these ultimate cooling machines, and things get stranger still, with heat flowing backwards and temperature itself ceasing to have any meaning.
Advertisement
The question of what it means for something to cool is thousands of years old. In 450 BC, the Greek philosopher Parmenides thought up frigidum primum, a hypothetical substance that was as cold as possible and could imbue other objects with coldness. In 1664, after a rigorous study of candidate substances like air or water, the chemist Robert Boyle declared the idea unjustifiable. He and his contemporaries instead believed that heat and cold were separate entities produced through chemistry or by living things.
Today, we understand that hot and cold are relative properties of a substance determined by how much its constituent atoms are jiggling around. If those atoms have lots of energy and are moving fast, it will be hotter than if they have less energy and are moving more slowly.
What is the coldest temperature?
When it comes to how cold things can get there is a fundamental limit. In 1848, Lord Kelvin introduced the idea of absolute zero. This is the point at which the atoms in an object stop moving entirely and so can’t get any colder. By extrapolating from experiments, Kelvin calculated that this point would be reached at -273°C. (In case you are wondering if there is a hottest possible temperature, there are physical limits – atoms can’t move faster than the speed of light, for instance – but the specific number isn’t known.)
In the decades that followed, Kelvin and others developed thermodynamics, the theory that relates heat, work and a quantity called entropy. Entropy is a measure of disorder. When atoms are zipping about chaotically, they are said to have high entropy, and when they are moving more slowly or just vibrating gently in neat patterns, they have lower entropy. We soon learned that heat always flows from warm to cold so that objects naturally reach equilibrium with their environment – think of a cool drink warming up in the summer sun – unless we put in some energy to push things the other way.
Freezers use energy to cool things down and high-end machines can usually only push this as far as -80°C (-112°F). But there is a relatively simple way to achieve a far lower temperature – pressurise some helium gas and expel it through a nozzle to create liquid helium at a temperature of -269°C, just 4 degrees above absolute zero. Immerse an object in this liquid and it will get very cold indeed.
How laser cooling works
In the late 1980s, three physicists independently devised a way of getting even closer to absolute zero, and it would prove to have huge consequences. The researchers in question were at the école Normal Supérieure in Paris, , then at AT&T Bell Laboratories in New Jersey, and at the National Institute for Standards and Technology in Maryland.
They started with a gas of atoms such as caesium or rubidium buzzing around in a vacuum chamber. The researchers then hit each atom with a laser pulse oriented the opposite way to the direction the particle was travelling in. By doing this many times, they could slow the atoms down to a hundred-millionth of their original speed, so that they were barely moving. And because heat equates to atomic movement, they were now very cold. The record was 40 microkelvin, just a few tens of millionths of a degree above absolute zero. In 1997, the researchers shared the Nobel prize in physics for inventing this cooling method.
In the early 1990s, three other physicists, and at the University of Colorado Boulder and, separately, Wolfgang Ketterle at the Massachusetts Institute of Technology, took yet another step down the temperature ladder. Their method, known as evaporative cooling, used lasers not just to slow atoms down, but also to shoot the warmest ones away, leaving only the coldest of the cold. They managed to cool rubidium and sodium atoms down to temperatures measured in nanokelvin, about a thousandth of the temperature achieved in the 80s. It won the trio a Nobel prize of their own in 2001.
All this allowed us to step through the door into the quantum world. It is a strange realm, where particles can sometimes exist in two places at once, or be entangled such that they can apparently influence each other instantaneously over vast distances. But these quantum behaviours can only be observed when the particles are undisturbed by outside influences, which includes even the vibrations of neighbouring atoms. With heat all but removed and the atoms now still, we could explore the quantum world freely.
One of the first things Ketterle and his colleagues discovered, in 1995, was a new state of matter that had been predicted by Albert Einstein 70 years before, but never observed. When you cool objects, you expect them to freeze into a solid. But hundreds of thousands of ultracold atoms can behave as a uniform chunk of quantum matter, known as a Bose-Einstein condensate. They lose their individual identity and slosh around as an odd liquid that can sometimes flow uphill. “As a general principle, if you can reach lower temperatures, you are bound to make discoveries,” says Ketterle. “If you live in a desert, you could not discover ice unless you build a refrigerator. That’s what we did when we reached nanokelvin temperatures.”
Since then, these cooling methods have created the booming field of ultracold physics, and we have learned how to cool molecules as well as atoms. It isn’t overstating it to say this has transformed physical science.
Take the way we design materials. Say we want to create a superconductor, a material that conducts electricity with no resistance. In the past, scientists would find a material – a metal perhaps – that did the job reasonably well, then investigate its atomic structure so they could better understand what makes an exceptional conductor. But that is a tricky job because the inner workings of materials are so complex.
Ultracold physics allows a better approach. Scientists can take molecules that are almost motionless and join them together like building blocks, assembling a material from scratch. This allows them to easily test hypotheses about how the internal structure of materials dictates their properties. This approach helped us grasp that a quantum property of atoms called spin seems to play an important role in some types of superconductor, for example.
Chemistry has also seen dramatic advances. Laser cooling can be used to slow down and steer chemical reactions so researchers can watch with unprecedented precision and clarity how molecules combine or split. In 2019, for instance, ‘s research group at Harvard University cooled potassium and rubidium atoms to 500 nanokelvin, forced them to start reacting, then used a laser to halt the reaction halfway through its natural course. What they spied when they hit pause was . By manipulating that frozen state with even more laser light, the researchers could effectively steer the rest of the reaction.
You might think that a billionth of a degree above absolute zero is as good as we can get. But lately researchers have been working on an entirely new cooling machine: the quantum fridge. This would be a system that employs the rules of the qauntum world to cool things down, and it could take us to a whole new frontier of cold – and weirdness.
Quantum fridges
Don’t picture a quantum fridge as remotely like a kitchen refrigerator – you won’t be putting your milk in one. Rather, they tend to consist of tiny quantum objects connected so that heat can be transferred from one to another. In 2019, at the National University of Singapore and his colleagues crafted such a quantum fridge out of three laser-cooled ytterbium atoms. “We could build a tiny machine, operate it in the quantum regime and see what happens,” he says. They experimented with putting these atoms in quantum states, for example quantum entangling them. Some of these effects allowed the fridge to by the rules of thermodynamics.
In the same year, and his colleagues at the University of Stuttgart in Germany made their own quantum fridge. Their design used molecules acting like the quantum bits, or qubits, in a quantum computer. The qubits were quantum correlated, meaning they could share information in a special way. Under these circumstances, the researchers showed that heat could flow the “wrong” way, so that a warm area would steal heat from a cooler area. The difference between the areas was minute, tens of billionths of a degree, but it was still a flagrant violation of the laws of thermodynamics.
For these quantum fridges, then, the normal rules clearly don’t apply. In a sense, this is no surprise. Thermodynamics is a theory designed to deal with sizeable collections of particles. Entropy, or disorder – that property so intertwined with temperature – only makes sense for a group of particles. When we are dealing with a quantum system of, say, two or three almost motionless particles, it is far less clear what “entropy” or “temperature” means.
Ronnie Kosloff at the Hebrew University of Jerusalem thinks thermodynamics should really apply in all circumstances. But perhaps, he says, we need a new version of the theory –quantum thermodynamics – that uses a definition of temperature that works at a quantum level; then the behaviour of quantum fridges might make more sense. One promising option is to use information as a proxy for temperature. It is another quantity that is closely related to entropy, but one we understand fairly well at a quantum level.
Quantum thermodynamics remains a work in progress. And quantum fridges are still being refined. But one thing is becoming abundantly clear: when we reach the bottom of the temperature scale, heat and cold become far more slippery concepts.
