In the Münster–Karlsruhe proof-of-concept device for an electrically driven, chip-scale single-photon emitter, a single carbon nanotube (center) is placed across a nanophotonic waveguide, and tied to gold electrodes. When a bias is applied to the electrodes, the tube pumps out single photons. Superconducting nanowires on either side of the nanotube were used as single-photon detectors to test the system. [Image: W. Pernice/WWU]
The past few years have seen considerable progress on the long road to quantum computers, though much remains to be done (see “Quantum computing: How close are we?,” OPN, October 2016). One persistent stumbling block toward all-optical versions of quantum computing has been getting reliable single-photon emitters that can be integrated at chip scale.
A team led by scientists from the University of Münster and the Karlsruhe Institute of Technology, Germany, has now proposed a possible solution: drop a carbon nanotube into the mix (Nat. Photon., doi: 10.1038/nphoton.2016.178). While their device is very much a proof of concept, the scientists believe that the use of such nanotubes—which can be electrically driven to emit photons, rather than requiring a pump laser source—could “pave the way toward all-photonic quantum circuits on a chip.”
Chip-scale conundrum
Many recent, promising developments in quantum computing have revolved around using the spins of atomic nuclei or electrons, arrays of trapped ions, or other matter systems as quantum bits or “qubits.” All-photonic systems have been a tougher sell. That’s partly because of the lack of interaction between photons, which creates difficulties in creating all-photonic logic gates—but also because of the engineering challenge implicit in getting single-photon sources miniaturized to the scale of a computer chip.
One problem with such scaling, according to the Münster–Karlsruhe team, is that most single-photon emitters tend to be optically pumped. Those systems require highly evolved, precise and space-consuming optical filters to remove the pump light, which would otherwise swamp the single-photon signal. And, because those filters need to be tied closely to the emitter wavelength, they can substantially restrict the ability for tuning of the single-photon characteristics.
Nanotubes to the rescue
The Münster–Karlsruhe team, along with scientists from the Moscow State Pedagogical University, Russia, sought to overcome these difficulties by turning to a single-photon emitter that’s electrically driven rather than optically pumped. Their emitter of choice: semiconducting single-walled carbon nanotubes (sc-SWCNTs), hollow carbon cylinders that are only a few microns in length and a mere nanometer or so in diameter.
A variety of recent experiments have shown sc-SWCNTs can emit light in both the visible and infrared bands when electrically or optically stimulated, and that they can act as single-photon sources under optical pumping. But whether the tubes could churn out reliable single photons under electrical triggering has remained an open question—the answer to which could open up some new avenues toward getting such emitters down to the chip scale.
To test the feasibility of the nanotube solution, the Münster–Karlsruhe team fashioned a proof-of-concept device that began with a photonic integrated circuit consisting of a single nanophotonic waveguide etched into a silicon chip. Across this waveguide, they placed a single carbon nanotube, flanked on either side along the waveguide by two superconducting nanowire single-photon detectors (SNSPDs) to pick up the signal of the emitted photons. All three elements—the nanotube and the two nanowires—were attached to gold electrodes, to provide a system that is completely electrically driven.
Antibunching signal
Upon flipping the proverbial switch and applying an electrical bias to the nanotube, the tube began emitting photons at room temperature, as predicted. The researchers then cooled the system down to around 1.6 K using liquid helium and studied the signal from the SNSPDs. They found a “pronounced” signature in the data of so-called antibunching—a type of sub-Poisson photon statistics that constitutes a key test of single-photon emission.
The Münster–Karlsruhe group points to a number of advantages for the nanotube-based scheme. One, as mentioned, is that using nanotubes eliminates the need for pump lasers and bulky optical filtering components. Also, the emission wavelength—varying from the visible to telecom-relevant infrared bands—hinges on the chirality and diameter of the nanotube, which can be selected relatively straightforwardly in a sorting step during fabrication. That could make the approach more tunable, and applicable to different uses, than some alternatives that depend on optical pumping.
The research team is quick to stress that the work represents “fundamental research,” rather than a practical device; for one thing, the key experiments took place at cryogenic temperatures. Yet the researchers believe that the system may, in the words of team leader Ralph Krupke of Karlsruhe, “overcome a limiting factor” in efforts toward on-chip quantum photonic circuits.