Researchers at Universität Innsbruck used a linear ion trap (Paul trap) to levitate a single nanoparticle—and, by interfering scattered light from the particle with its mirror image, achieved dramatic gains in the ability to detect the particle’s position. [Image: Quantum Interface Group, Universität Innsbruck] [Enlarge image]
In experiments probing the quantum mechanics of nanoscale objects, levitated optomechanics—the art of trapping, suspending and cooling tiny particles with light—is a fundamental part of the toolkit. But while a few recent studies have used these setups to cool nanoparticles to their motional ground state, there’s still the possibility of light-induced decoherence that can gum up the works in sensitive quantum-mechanical tests.
One way to get around that hurdle is to suspend the particle using an electrical ion trap, such as a Paul trap, or a magnetic trap, which offer environments free of decoherence induced by stray photon scattering. But the detailed geometry of ion traps, which limits optical access to the particle for ultra-precise position sensing, puts the quantum regime out of reach in those setups.
Now, a team led by Tracy Northup at Universität Innsbruck, Austria, has fashioned a clever self-interference detection method that sidesteps these limitations (Phys. Rev. Lett., doi: 10.1103/PhysRevLett.129.013601). The researchers say the technique not only overcomes the obstacles to doing macroscopic quantum experiments in low-noise Paul traps, but also gets around some previous fundamental limitations on motion detection and quantum control in all levitated-nanoparticle experiments.
The result, the team believes, could open up new avenues for pushing macroscopic objects close to their quantum ground state, and for using levitated nanoparticles in quantum sensors.
Whatever the platform, a typical levitated-optomechanics experiment relies on ultra-precise measurement of the suspended particle’s position. In most experiments, this is achieved by shining laser light on the suspended particle, and interfering the scattered light from the particle with another, reference laser beam. Changes in the interference pattern allow the particle’s motion to be reconstructed—and controlled by electrostatic or laser cooling.
Unfortunately, this measurement setup has its limitations. The biggest one is the mode mismatch between the dipole radiation of the scattered light from the particle and the Gaussian profile of the reference laser. This reduces the visibility of the interferometric fringes, and thus the sensitivity of the position measurement. In addition, the conventional setup gathers only half of the light scattered by the particle for interference with the reference beam, discarding potentially valuable additional information on the particle’s position.
Northup’s group sought a way to overcome these limitations—and one that could be implemented in a linear Paul trap. To do so, the team built on previous work from the team of another Innsbruck researcher, Rainer Blatt, involving the levitation of individual ions, and extended it to the nanoparticle scale.
In the basic setup, one side of scattered light from an illuminating laser beam is sent through a collimating lens to an avalanche photodiode (APD), and scattered light from the other side is sent through a similar lens to a flat mirror, with the reflected signal also collected at the APD. Interfering the two signals allows the particle’s position to be detected with exquisite precision. [Image: L. Dania et al., Phys. Ref. Lett., doi: 10.1103/PhysRevLett.129.013601 (2022); CC-BY 4.0] [Enlarge image]
While these experiments constitute a tour de force of laboratory measurement, the principle behind them is simplicity itself. Instead of interfering scattered light from the suspended particle with a Gaussian laser beam, the light scattered from one side of the particle is interfered with light scattered from the other side, reflected off of a distant, planar mirror. The phase of the interference signal depends on the particle–mirror distance, and thus helps pin down the particle position.
The self-interference setup nicely cleans up both of the ragged edges of the conventional position measurement using a reference laser. Because one scattered dipole field is being interfered against another, there’s no mode mismatch. And the setup uses scattered light from both sides of the particle, rather than only one side—thereby capturing more detailed information on the particle’s position.
Stunning SNR improvement
Northup and her colleagues found a way to put this approach into effect on a nanoparticle suspended in a linear Paul trap. They started by confining a 300-nm-diameter silica sphere in the trap at high vacuum, and then implementing an initial feedback cooling step using the conventional method of forward-scattering detection and interference. This cooled the particle to the millikelvin range, bringing its amplitude of motion below an optical wavelength. At that point, the team switched on its self-interference detection approach, interfering 780-nm scattered laser light from one side of the suspended particle against a mirror image of the light scattered from the other side.
The team found that its self-interference method enabled a stunning 38-dB improvement in signal-to-noise ratio compared with the conventional method of interference against a reference laser beam. What’s more, using the self-interference signal as an input, the team was able to implement further feedback cooling to as low as 18 mK. That’s far below the minimum temperatures achievable in such a trap using conventional forward-detection feedback.
Expanded palette for quantum experiments
The researchers report that the interferometric visibility between the particle and its mirror image reached 70% in their experiments. That’s not bright enough to take a particle to its motional ground state. However, the team believes that it can reach visibilities of nearly 100% by including a hemispheric mirror inside the vacuum chamber, rather than using a planar mirror outside of it.
This step, Northup and her colleagues believe, should allow the self-interference method to overcome the previous obstacles to doing quantum-optics experiments in Paul traps. “Given the low intrinsic decoherence in these traps due to the lack of photon scattering,” the authors write, “we expect that this detection approach will enable quantum experiments with macroscopic objects with a unique degree of isolation.” And, the team stresses, the detection technique’s advantages in sensitivity and signal-to-noise, while demonstrated in a Paul trap, can be extended “to setups that confine dipolar emitters by magnetic or optical forces.”