Calculating the properties of a quantum particle in a box is something most physics students have to do as part of their degree course – but actually creating such a simple system in the lab can be an experimental challenge. Now, however, physicists in the UK are the first to create a Bose–Einstein condensate (BEC) in a 3D optical-box trap, which resembles a tin can. The breakthrough could allow physicists to study a range of multi-body physics phenomena in controlled conditions.

The first BEC was made in 1995 in Nobel-prize-winning work that involved cooling a cloud of rubidium-87 atoms down to temperatures of near absolute zero. The atoms settle into a quantum state that extends over a macroscopic volume, which means that the BEC behaves like a superfluid. In addition to being fascinating in their own right as a new state of matter, BECs are interesting because they are created under very controlled conditions, which allows them to be manipulated to resemble a variety of quantum phenomena.

Earlier this year, for example, physicists carried out an experiment in which a BEC behaved much like a Josephson junction – a device that is normally make from a superconductor. As such, BECs can be used as "quantum simulators" to gain a better understanding of less-accessible quantum systems, ranging from magnets to superconductors and neutron stars. Unfortunately, physicists have so far only been able to create BECs in traps where the trapping potential – and so the atomic density – varies harmonically, which is no good for anyone simulating, say, electrons in solids, as these systems tend to have homogenous particle densities.

Lids on a tin can

Now, however, Zoran Hadzibabic and colleagues at the University of Cambridge are the first to create a BEC in a 3D trap that – for most practical purposes – has a constant potential in all three directions. "Our trap is an optical box made out of green light: a dark region of empty space is surrounded by thin walls of light that repel the atoms and keep them confined inside the box," says Hadzibabic.

His team created the box by imprinting a phase pattern onto a conventional laser beam. The result is a hollow tube of green laser light and two "sheet-shaped" laser beams lying perpendicular to the tube. "The beams form the end lids that close up our optical tin can," says Hadzibabic. The effects of gravity, which would otherwise distort the box, were eliminated by suspending the atoms using a magnetic field.

Before creating the BEC, Hadzibabic's team had to cool down a cloud of rubidium-87 atoms to nanokelvin temperatures. This involved first gradually lowering the harmonic trapping potential so that faster-moving hot atoms escape the trap, leaving only cooler atoms behind. As it undergoes this process of "evaporative cooling", the cloud shrinks until it is small enough to fit inside the optical box. The box is then switched on and the harmonic trap is turned off slowly.

The next step involves subjecting the atoms to a final round of evaporative cooling to get the gas to a temperature of below about 90 nK, where it turns into a BEC. This was done by adjusting the intensity of the laser light that creates the walls of the trap. Faster-moving hot atoms were able to penetrate the walls and exit the trap, while the cooler atoms cannot – causing evaporative cooling.

Einstein's right again

To confirm that they actually had a BEC in a uniform potential, the researchers turned off the box trap and let the gas expand freely while measuring the velocity distribution of the atoms. A large peak at very low velocity confirmed that a BEC had formed and the shape of the peak contained information about the shape of the box trap. The velocity distribution also revealed the temperature below which the atoms condensed into a BEC. This temperature was first predicted by Einstein in 1925 and Hadzibabic says their analysis is the best experimental confirmation so far.

While the box trap is a good approximation to a constant potential in 3D, it is not perfect, although Hadzibabic argues that it is good enough for most applications. "In our trap," he says, "we can estimate that more than 80% of the atoms live within the region where the density deviates by less than 10% from the average value, so these atoms should heavily dominate all experimental signals." In contrast, less than 20% of the atoms in a conventional harmonic trap lie in a region that is representative of the average density.

Focus on phase transitions

Now that they have created a near-homogenous BEC, the researhcers are keen to use it to simulate a range of quantum systems. In particular, the set-up should be good for studying how a system makes the phase transition from a cold gas to a BEC. The team’s first target is to study the effects of inter-particle interactions on Bose–Einstein condensation of a homogeneous gas or fluid. This problem was first proposed in 1957 by Chen Ning Yang and Tsung-Dao Lee – Chinese-American physicists who also won the Nobel prize that year for unrelated work on particle physics.

"This problem has been studied in liquid helium, but many questions remain open and the agreement between theory and experiment has not been reached," explains Hadzibabic. The team is also looking at doing other experiments in which the interactions between atoms can be fine-tuned. This will involve modifying the experiment to use potassium-39, which is more difficult to trap but better for creating tuneable interactions.

The results are described in Physical Review Letters and a preprint is available on arXiv.