QTu2B.5.pdf Research in Optical Sciences © OSA 2014 Detection of Single Photons Using Slow light Zhifan Zhou 1,* , Zhongzhong Qin 1 , Yami Fang 1 , Ryan T. Glasser 2 , Ulrich Vogl 3 , Jietai Jing 1 , Weiping Zhang 1 1 Quantum Institute for Light and Atoms, State Key Laboratory of Precision Spectroscopy, Department of Physics, East China Normal University, Shanghai 200062, People’s Republic of China, 2 Quantum Measurement Division, National Institute of Standards and Technology and Joint Quantum Institute, NIST and University of Maryland, Gaithersburg, Maryland 20899, USA, 3 Max Planck Institute for the Science of Light, Günther-Scharowsky-Straße 1, Bau 24, 91058 Erlangen, Germany. *Author e-mail address: zfzhoucom@gmail.com Abstract: We demonstrate photodetection with single-photon level sensitivity using slow light through a four-wave mixing process with high dispersion and high gain in hot rubidium vapor. The delay times are dependent on the few-photon input states. OCIS codes: (190.4380) Nonlinear optics, four-wave mixing; (270.0270) Quantum optics; (230.0230) Optical devices. 1. Introduction Single-photon detection [1, 2] is crucial for a variety of quantum-information and measuring tasks. Substantial progress has been made by use of the photoelectric effect and the photothermal effect at the single-photon level [1, 2]. Alternatively, atomic-vapor-based single-photon detection [3, 4] has been suggested to achieve ideal detection efficiency in principle by converting a single photon into many photons through quantum light-atom interfaces [5, 6]. By employing the high dispersion within the conical emission regime in a four-wave mixing (FWM) process in hot rubidium vapor, we can observe ultraslow matched pulse propagation even when injected with single-photon level optical pulses. Due to the high gain associated with this regime, the single-photon level injection can be amplified up to a macroscopic level, enabling easy detection with conventional detectors (i.e. non-single-photon detectors). Furthermore, we experimentally show that the delays depend strongly on the average input few-photon-number of the injected probe [7], which paves the way to all-optical photon-number-resolving detection. 2. Experimental setup Fig.1. (A) Double-lambda energy level scheme and (B) a schematic diagram of experimental setup. We generate the conical emission through a FWM process in hot rubidium vapor with a double-lambda scheme. The double-lambda scheme and the experimental setup are shown in Figure 1. The double-lambda scheme (Fig.1A) allows for conversion of pump photons into probe and conjugate photons stimulated by spontaneous emission. The steep dispersion based on the double-lambda scheme is important for phase matching in the non-degenerate parametric process since it can eliminate phase mismatch due to the different group indices the probe and conjugate experience. By increasing the pump power and the atomic density, the strong nonlinear coupling of the pump and atoms provides large gain for the spontaneously emitted photons in a single-pass configuration, resulting in an output pattern of a ring. The conical emission due to the non-degenerate four-wave mixing is comprised of photons at the Stokes and Anti-Stokes frequencies in the same spatial modes (Fig.1B).