In our recent paper [1], we reported observations of photon blockade by one atom strongly coupled to an optical cavity. In support of these measurements, here we provide an expanded discussion of the general phenomenology of photon blockade as well as of the theoretical model and results that were presented in Ref. [1]. We describe the general condition for photon blockade in terms of the transmission coefficients for photon number states. For the atom-cavity system of Ref.[1], we present the model Hamiltonian and examine the relationship of the eigenvalues to the predicted intensity correlation function. We explore the effect of different driving mechanisms on the photon statistics. We also present additional corrections to the model to describe cavity birefringence and ac-Stark shifts.

Quantum information science attempts to exploit capabilities from the quantum realm to accomplish tasks that are otherwise impossible in the classical domain. Although sufficient conditions have been formulated for the physical resources required to achieve quantum computation and communication, there is a growing understanding of the power of quantum measurement combined with the conditional evolution of quantum states for accomplishing diverse tasks in quantum information science. For example, a protocol has recently been developed for the realization of scalable long-distance quantum communication and the distribution of entanglement over quantum networks. Here we report the first enabling step in the realization of this protocol, namely the observation of quantum correlations for photon pairs generated in the collective emission from an atomic ensemble. The nonclassical character of the fields is demonstrated by the violation of an inequality involving their normalized correlation functions. Compared to previous investigations of non-classical correlations for photon pairs produced in atomic cascades and in parametric down-conversion, our experiment is distinct in that the correlated photons are separated by a programmable time interval (of about 400 nanoseconds in our initial experiments).

EXPERIMENTAL DETAILSOur experimental setup is depicted by the simple drawing in Fig. 1A of the manuscript, with many of the technical aspects described in more detail in Refs.[S1, S2]. After releasing a cloud of atoms from a magnetooptical trap (MOT) above the cavity, transverse cooling beams illuminate the cavity region, at which point an atom can be loaded into the intracavity far-off resonance trap (FORT), which is matched to a standingwave, TEM 00 mode along the cavity axis. The trap depth is U 0 /k B = 2.3 mK (47 MHz), and because its wavelength is λ F = 935.6 nm, the potential for the atomic center-of-mass motion is only weakly dependent on the atom's internal state [S2]. The cavity length is actively stabilized with an auxiliary laser at wavelength λ C = 835.8 nm that does not interfere with the trapping or the cQED interactions. Relevant cavity parameters are length l 0 = 42.2 µm, waist w 0 = 23.6 µm, and finesse F = 4.2 × 10 5 at 852 nm. For our system, the Rabi frequency 2g 0 for a single quantum of excitation is given by g 0 /2π = 16 MHz, where g 0 is based upon the reduced dipole moment for the 6S 1/2 , F = 4 ↔ 6P 3/2 , F = 3 transition in atomic Cs (Fig 1B). The amplitude decay rates (κ, γ) due to cavity losses and atomic spontaneous emission are κ/2π = 4.2 MHz, and γ/2π = 2.6 MHz. Since g 0 (κ, γ), strong coupling is achieved, resulting in critical photon and atom numbers n 0 ≡ γ 2 /(2gWith an atom loaded into the intracavity FORT, our protocol for the generation of single-photon pulses consists in illuminating the atom with a sequence of laser pulses according to the timing diagram shown in Fig. 1(c) of the manuscript. Within each trial, the first pulse Ω 3 (t) contains light tuned 10 MHz blue of F = 3 → F = 3 , which initiates the adiabatic transfer F = 3 → 4 between the ground hyperfine levels, with the emission of a photon into the cavity mode. This transformation is principally accomplished via "dark" eigenstates of the atom-cavity system, with no contribution from the excited level F = 3 , and hence with a concomitant reduction of fluorescent loss [S3, S4, S5]. The second pulse Ω 4 (t) is tuned 17 MHz blue of F = 4 → F = 4 and recycles the atom back to the F = 3 ground state through spontaneous decay F = 4 → F = 3. Each Ω 3,4 field consists of two orthogonal pairs of counter-propagating beams in a σ + −σ − configuration. The detuning between the 3 → 4 transition at ω 43 and the cavity resonance ω C is ∆ CA ≡ ω C − ω 43 = 2π × 9 MHz [S6].We now provide some additional details on the optical path from the cavity to the detectors. After emerging from the vacuum chamber window, the path includes a polarizing beam splitter (PBS), several dichroic mirrors and two interference filters. The light is next coupled into a single-mode fiber, and then split using a 50/50 fiber coupler. The two output fibers of the coupler are connected to fiber-coupled avalanche photodiodes (APD), labelled D A and D B . LOSSES AND EFFICIENCIESPhotons generated in the cavity are subject to various types of loss along thei...

No abstract

On the occasion of the hundredth anniversary of Albert Einstein's annus mirabilis, we reflect on the development and current state of research in cavity quantum electrodynamics in the optical domain. Cavity QED is a field which undeniably traces its origins to Einstein's seminal work on the statistical theory of light and the nature of its quantized interaction with matter. In this paper, we emphasize the development of techniques for the confinement of atoms strongly coupled to high-finesse resonators and the experiments which these techniques enable.(Some figures in this article are in colour only in the electronic version) From Einstein to cavity QEDIn the years prior to his seminal 1905 papers, Albert Einstein had given much thought to the statistical properties of electromagnetic fields [1], especially with regard to the theory of black-body radiation developed by Max Planck [2]. Einstein realized that the quantization of light-particularly the creation and annihilation of 'light quanta'-is something more fundamental than a tacit consequence of the assumption that the total energy of a black-body is discretely distributed between a set of microstates. Beginning in 1905 with On a heuristic point of view about the creation and conversion of light [3] and in four subsequent papers on quantization [4][5][6][7], he laid the foundations of the 'old quantum theory ' [8], summarized in what is commonly referred to as the 'light quantization hypothesis':. . . the energy of a light ray emitted from a point [is] not continuously distributed over an ever increasing space, but consists of a finite number of energy quanta which are localized at points in space, which move without dividing, and which can only be produced and absorbed as complete units [3].