Strong coupling without touching

If an excited atom is put in between high-reflectance mirrors, the light emitted by the atom can be bounced back and be re-absorbed and be re-emitted, again and again until the light leaks out of the mirrors or is re-emitted into a wrong direction. This coherent oscillation between atom and light, called vacuum Rabi oscillation, results from interference between two superposition states of the atom and the photon, which have different energies or frequencies, or vacuum Rabi splitting. The vacuum Rabi oscillation and splitting demonstrate the quantum nature of light. They also provide single-photon non-linearity for photonic applications. In recent years, these effects have been considered particularly useful for quantum information processing [1] since the oscillation presents a natural mechanism for information exchange between a stationary quantum bit carried by the atom and a flying quantum bit carried by the photon, making scalable and distributed quantum computing possible. Realization of the vacuum Rabi oscillation and splitting, however, is highly non-trivial. It requires the so-called strong coupling condition, that is, the exchange between the atom and the photon must be faster than the photon leaks out of the space enclosed by the mirrors, i.e. the cavity, and faster than the atom emits the light into a wrong direction. The technique to realize the strong coupling condition is to make the cavity small so that the photon is bounced back more frequently to the atom and to make the mirrors more reflective so that the photon can be bounced more times. That means to decrease the cavity volume V and to increase the quality factor Q. Remarkable progresses have been made to make optical microcavities with small V and high Q, facilitating observation of vacuum Rabi oscillation and splitting in highly dissipative solid-state systems [2,3]. However, there is no systematic method to reduce V/Q—the key factor for realizing strong coupling (see Fig. 1a). The reason appears to be fundamental. When the size of a cavity is made as small as comparable to the light wavelength, small deformation of the cavity would result in large evanescent wave and hence photon leakage [4]. (a) An atom and a small cavity have strong coupling but the cavity leakage is also strong. A large cavity has small photon leakage but its coupling to the atom is small (so emission goes to free space). The atom and the large cavity can have strong coupling, with photons exchanged through but not occupying the small cavity—using the quantum tunneling effect. (b) In the classical world, two lower players have to provide sufficient energy to pass the ball through the higher player. (c) In the quantum world, the two lower players can pass balls through the wall via the higher players but with energy of the balls insufficient to reach the higher player. A recent proposal by [5] provides a general solution, using a divide-and-conquer strategy. A small cavity with relatively low Q is used to enhance the interaction between the emitter and the photon, while the photon stays in a large cavity with high Q. [5] demonstrate the strong coupling by showing Rabi splitting, Rabi oscillation, and single-photon non-linearity of the system. The scheme at the first sight does not seem to work since the emitter has no contact with the high-Q large cavity and if the photon enters the small cavity, it would leak out. The clever point of the scheme is: in the quantum mechanical world, the photon can interact with the emitter without occupying the small cavity! Similar to the ball game in Fig. 1b, a classical particle needs to have sufficient energy to pass a barrier. The small cavity is designed to have a frequency higher than those of the emitter and the large cavity, which would prevent the emitted photons entering the large cavity. However, quantum ball games allow tunneling of a barrier (Fig. 1c). So the emitter and the large cavity can indeed exchange photons. The small cavity would not be occupied since its frequency is not right, which prevents photon leakage. It looks the scheme applies generally to all types of photonic systems—photonic crystals, microspheres, microtoroids, etc. Considering the simplicity of the scheme, we expect applications be found soon.