Portland, Ore. - Recent developments bode well for comprehensive quantum information systems. Work at labs in Colorado and Austria has increased the ability to store quantum states on groups of atoms, and researchers in Georgia have found a way to transfer quantum states over networks.
Two groups-at the National Institute of Standards and Technology (NIST; Boulder, Colo.) and at the University of Innsbruck's Institute of Theoretical Physics-have separately pushed up the number of ions that can exist in a simultaneous superposition of states. Just as NIST researchers were reporting a successful experiment observing six rubidium ions in a synchronized state of superposition (see Nov. 28, page 12), the Innsbruck group announced the observation of eight calcium ions in a magnetic ion trap. Previously, quantum computing research had established quantum entanglement in five photons.
Meanwhile, a team at the Georgia Institute of Technology has found a way to build quantum state "repeaters"-systems that regenerate a quantum state being transmitted over a network-which would enable larger quantum networks to be built. The first application of the development will likely be in emerging quantum encryption systems that operate over optical networks. In theory, quantum repeaters could exchange secure encryption keys. By demonstrating the temporary storage and retrieval of quantum information from a cloud of rubidium atoms, the Georgia Tech researchers have verified the possibility of building such systems.
"We have demonstrated that we can store quantum information on clouds of rubidium atoms for up to 10 microseconds, then read it back out," said postdoctoral researcher Thierry Chaneliere, working in the lab of professors Alex Kuzmich and Brian Kennedy at Georgia Tech. "This could someday enable a quantum repeater-one more building block for quantum computers and networks."
Qubits are the basic unit of quantum information, just as a bit, represented by a device with two possible states, is the basic unit of today's information processors. The essential difference between the two building blocks is the ability of qubits to represent two quantum states that can simultaneously exist on a single photon or ion. Reading the value of a quantum state, however, destroys the dual-state representation, so quantum computers and networks must be able to process states internally without interacting with the environment.
The disturbance of quantum states following a data read is what makes quantum encryption the ultimate secure protocol. Any attempt to eavesdrop on a stream of quantum states will alter them, making it possible for sender and receiver to detect the interception of information.
"We have made an important step forward, but it's still a building block. There will be a lot of steps and several more years before these things mature in a practical way," said Kuzmich. Also contributing to the work were doctoral candidates Dzmitry Matsukevich, Stewart Jenkins and Shau-Yu Lan.
Last year, Kuzmich's group reported transferring atomic-state information from rubidium atoms to photons, saying it was the first time quantum information had been transferred from matter to light. Now, in the new demonstration, the quantum information changed from photons to stationary atoms, stored from 500 nanoseconds to 10 microseconds, with the photon then regenerated with its original quantum information intact.
The experiment used two clouds of rubidium atoms at opposite ends of a 100-meter-long optical fiber that had been cooled to near absolute zero, thereby limiting the available quantum states. Both clouds were held captive by strong magneto-optical traps; then the first rubidium cloud was stimulated to emit a photon into the optical fiber, sending it to the second rubidium cloud. The photon contained the quantum information describing the resonance state of the rubidium atoms, and when it hit the second rubidium cloud the quantum information was transferred, under the direction of a control laser. The second cloud stored the photon for 500 nanoseconds to 10 microseconds before being induced to give it up, directly re-encoding the quantum information onto it.
The researchers characterized the mechanism by which the quantum information was transferred from the photon's spin to the atom's vibration as a light field excitation called a dark-state polariton, which can be later recovered from the atoms by inducing them to emit a photon.
"We store the information from the photon in the state of excitation of many atoms in the second ensemble," said Chaneliere. "It's really information about spin, but we store it in each atom in the ensemble, all of which are slightly flipped."
By storing quantum information-transferring the spin of photons to the vibrations of an atom and back again-Kuzmich's group has demonstrated the feasibility of a quantum repeater, which would enable future quantum communications to move beyond their current direct-connection limitation. Such systems would have virtually no distance limitation, since another repeater could always be added. Quantum registers and other computer memory components might also be enabled someday by the Georgia Tech store-and-retrieve approach.
"Now that we can store and retrieve quantum information, we want to work toward quantum networks where each node is a quantum computer," said Chaneliere. The lab recently demonstrated entanglement of two atomic qubits that were separated by a distance of 5.5 meters. Entanglement-a special kind of quantum synchronicity whereby reading out the state of one of an entangled pair determines the value of the other-had previously been demonstrated only over a distance measured in millimeters.
"We generated entanglement of atomic qubits and showed that we can take this entanglement and map it from atoms to photons," said Kuzmich.
As other groups, such as the NIST and University of Innsbruck researchers, push up the number of ions that can store information, the Georgia Tech demo of networked quantum states becomes a powerful method for linking qubits stored in different locations. Only eight ions in a superposition of states can potentially represent more than 65,000 quantum states, so modest increases in the physical size of quantum computing components can have an exponential impact on data throughput.
But maintaining the state of entanglement that allows a group of particles to represent large amounts of data simultaneously is difficult. In the case of particles like ions, near-absolute-zero temperatures are required to damp thermal vibrations, and complicated laser and magnetic field configurations are needed to maintain the quantum states.
"It is very difficult to control six ions precisely for a long enough time to do an experiment like this," said NIST researcher Dietrich Leibfried. The experimental conditions robust, however, with the results being repeated thousands of times.
In addition, the NIST system was in an extreme configuration in which all six atoms are in a spin-up and spin-down state simultaneously. In such a configuration, measuring the state of one ion would cause all six to collapse into specific states, loosing coherence.
For the Innsbruck experiment, the ions were in a more-independent configuration in which some of the information could be extracted without destroying the quantum coherence of the system.
- Chappell Brown contributed to this report.