Krill Prize 2015
Controlling Single Photons
Photons – the particles of light – are one of the most fundamental and basic elements in nature, and play a central and crucial role in nearly all the fields of science and modern technology. Modern digital communication relies almost solely on photons guided in optical fibers, and lasers are an elementary, indispensable part of scientific experiments, medical treatments, industry and more. Yet surprisingly, our ability to control light, especially at the level of single photons, is incredibly limited. As an example consider the seemingly trivial and fundamental problem of separating two incoming photons that arrive in a single laser pulse. Surprisingly – this task has been beyond the ability of even cutting-edge optical labs. Other seemingly trivial tasks like generation of just one photon, or counting the number of photons in a pulse without destroying them, or creating a switch for photons that is controlled just by photons – have been a long-standing goal and a subject of intense research efforts worldwide in the last couple of decades. In my group we have set the goal to achieve exactly such control over single photons, and in particular to attain deterministic photon-photon interactions, namely to demonstrate the ability of one photon to control another photon. To do so we built the technology to make single photons interact strongly, and therefor deterministically, with a system that naturally changes its behavior in response to interaction with just one photon – a single atom. Specifically, in our lab we use laser-cooling of atoms to bring single 87Rb atoms to the vicinity of chip-based, fiber-coupled high-quality optical microresonators. These microresonators enable confinement of light to micron-sized volumes for a long time, thereby enhancing the electric field associated even with a single photon to the level that allows fast and deterministic interaction with a single atom. The attainment of such state-of-the-art technologies have led to Using this technology, and based on a theoretical work we performed that studied the fundamental properties of such single-photon-single-atom interactions (published in PRA 84,033854 (2011)), we have succeeded in achieving for the first time the goal of making one photon control another one without any other control fields. In this work (Science 345, 903 (2014)) we demonstrate a single-atom based all optical switch, which reflects or transmits a single photon depending on the “command” given by a previous single-photon pulse. In particular, this photonic device also accomplished, for the first time, the task described above, of separating two incoming photons, as even within the same pulse the first photon is reflected, and the second is transmitted. Based on this mechanism we then demonstrated for the first time deterministic single-photon extraction, (submitted for publication). This is the strongest possible single-photon nonlinearity, yet the prospect of photon-photon interactions bears tremendous significance in physics, far beyond the mere technological achievement of controlling light by light. Photon-photon interactions are inherently non-classical, and are an enabling building block for fundamental tasks in quantum optics (for example, for the generation of optical Schrodinger-cat states) and quantum information science (for example, the realization of optical quantum logic gates and more). In particular, the potential of our scheme lies in its compatibility with scalable photonic architectures. The device is operated only by the single-photon pulses, which are all in-fiber, identical, and routed to the output ports. This means that a routed target photon can serve as the control photon in the next device, or that the same control photon could activate a few devices. The scheme we demonstrated therefore provides a versatile, robust, and simple building block for a variety of all-optical photonic devices, from quantum memory through singlephoton add/drop filters, to photonic quantum gates, all of which being completely passive and compatible with scalable quantum networks.