Florian Fertig (left) and Wei Zhang at Ludwig Maximilian University of Munich are part of a research team that demonstrated entanglement of two widely spaced quantum memories. [Image: J. Greune]
Large-scale quantum networks offer the potential of unhackable communications and pooled processing of widely distributed quantum computers, among other things. But for such networks to see the light of day, scientists must first show how to provide entanglement between any two or more users on the network when users want it.
Researchers have now taken a step toward this objective by showing how frequency conversion enables an entangled link to be set up between two quantum memories separated by up to 33 km of optical fiber. The photons used to establish the link had wavelengths close to that of the telecom band (Nature, doi: 10.1038/s41586-022-04764-4).
Overcoming the loss
Entanglement is a particularly striking quantum phenomenon, involving a correlation between two particles such that a measurement of one instantaneously fixes the state of the other. This holds true no matter how far apart the particles are, implying that entanglement is, in principle, an ideal way of setting up long-distance links—both to guarantee the security of quantum key distribution and to send quantum information via teleportation.
A key challenge in realizing such links lies in overcoming the loss within optical fiber. That loss lowers the odds that any given photon will make it to the far end of a stretch of fiber as the fiber’s length increases, thereby imposing an upper limit on the rate at which quantum communications can be carried out across a given distance.
As Harald Weinfurter of Ludwig Maximilian University of Munich (LMU), Germany, and colleagues point out in the latest research, the wavelength of the photons used to set up the link can make a big difference in the rate of entanglement. At 780 nm, on average just one photon in ten would make it to the far end of a 2.5 km length of fiber, while at close to the 1550 nm telecom sweet spot, that level of attenuation would only occur after about 50 km of fiber transmission.
High fidelity
To show the promise of frequency conversion for entanglement generation, Weinfurter and co-workers set up two quantum memories spaced some 400 m apart in separate buildings on the LMU campus. The researchers made each memory from a single atom of rubidium-87 and exposed the atoms to nanosecond-length laser pulses such that they emitted single photons at 780 nm as they spontaneously dropped back down to their ground states. Conservation of angular momentum dictated that the polarization of the emitted photons became entangled with the spin of the respective atoms.
The conversion was mastered by Christoph Becher and colleagues at Saarland University, Germany. It involved mixing the 780-nm photons with pump light at 1607 nm inside a nonlinear crystal, such that the difference in frequencies yielded photons at 1517 nm. Their polarizations preserved, these particles were then sent along a fiber to a central station where a Bell state measurement swapped entanglement from the atom-photon pairs to the two atoms, with successful entanglement being “heralded” by a signal sent back to the two memory nodes.
By adding spooled fiber into the photons’ path, the researchers were able to vary the path length between 6 km and 33 km. They found that, across all distances, the fidelity remained above 0.5—indicating that they were able to generate entangled states in each case—even though the number dropped off at longer distances (from 0.83 at 6 km to 0.62 at 33 km).
“Our results indicate that quantum frequency converters could be used to construct large-scale quantum networks, and our experimental setup provides a platform to implement secure quantum communication protocols,” Tim van Leent and Pooja Malik at LMU write in the research briefing accompanying their paper.
Solving remaining problems
The researchers acknowledge that more work needs to be done before the scheme can be used in the real world. In particular, they point out that the setup currently suffers from very low efficiencies. The small odds of any given attempt at entanglement being successful and the finite time needed for heralding meant that they were able to successfully generate entangled pairs of atoms over 33 km only once every 85 seconds on average. Given that the quantum memories typically remained coherent for less than a millisecond, they were able to supply entangled states for no more than about 1 in 105 of the time.
As they note, this is a particular problem when it comes to realizing quantum repeaters—devices that extend the distance between communicating parties by establishing multiple individual links and then swapping the entanglement to progressively greater distances.
The researchers suggest a number of potential routes for improving the efficiency. These include increasing the coherence time of the quantum memories, using optical cavities to improve the collection of photons emitted by the atoms and running arrays of single-atom memories to carry out multiple entanglement attempts simultaneously. They acknowledge that none of these solutions is likely to prove easy but say there are no fundamental barriers to overcome. Mastering the technology, says Becher, “is more a question of money and workforce.”