By harnessing aspects of quantum mechanics, communication and information processing could be radically transformed. Promising forms of quantum information technology include optical quantum cryptographic systems and computing using photons for quantum logic operations. As with current information processing systems, some form of memory will be required. Quantum repeaters, which are required for long distance quantum key distribution, require quantum optical memory as do deterministic logic gates for optical quantum computing. Here, we present results from a coherent optical memory based on warm rubidium vapour and show 87% efficient recall of light pulses, the highest efficiency measured to date for any coherent optical memory suitable for quantum information applications. We also show storage and recall of up to 20 pulses from our system. These results show that simple warm atomic vapour systems have clear potential as a platform for quantum memory.

The bandwidth and versatility of optical devices have revolutionized information technology systems and communication networks. Precise and arbitrary control of an optical field that preserves optical coherence is an important requisite for many proposed photonic technologies. For quantum information applications, a device that allows storage and on-demand retrieval of arbitrary quantum states of light would form an ideal quantum optical memory. Recently, significant progress has been made in implementing atomic quantum memories using electromagnetically induced transparency, photon echo spectroscopy, off-resonance Raman spectroscopy and other atom-light interaction processes. Single-photon and bright-optical-field storage with quantum states have both been successfully demonstrated. Here we present a coherent optical memory based on photon echoes induced through controlled reversible inhomogeneous broadening. Our scheme allows storage of multiple pulses of light within a chosen frequency bandwidth, and stored pulses can be recalled in arbitrary order with any chosen delay between each recalled pulse. Furthermore, pulses can be time-compressed, time-stretched or split into multiple smaller pulses and recalled in several pieces at chosen times. Although our experimental results are so far limited to classical light pulses, our technique should enable the construction of an optical random-access memory for time-bin quantum information, and have potential applications in quantum information processing.

Received Month X, XXXX; revised Month X, XXXX; accepted Month X, XXXX; posted Month X, XXXX (Doc. ID XXXXX); published Month X, XXXX The realization of quantum memory using warm atomic vapor cells is appealing because of their commercial availability and the perceived reduction in experimental complexity. In spite of the ambiguous results reported in the literature, we demonstrate that quantum memory can be implemented in a single cell with buffer gas using the geometry where the write and read beams are nearly co-propagating. The emitted Stokes and anti-Stokes photons display cross-correlation values greater than 2, characteristic of quantum states, for delay times up to 4 s.

We propose a photon echo quantum memory scheme using detuned Raman coupling to long lived ground states. In contrast to previous 3-level schemes based on controlled reversible inhomogeneous broadening that use sequences of π-pulses, the scheme does not require accurate control of the coupling dynamics to the ground states. We present a proof of principle experimental realisation of our proposal using rubidium atoms in a warm vapour cell. The Raman resonance line is broadened using a magnetic field that varies linearly along the direction of light propagation. Inverting the magnetic field gradient rephases the atomic dipoles and re-emits the light pulse in the forward direction. c 2017 Optical Society of America OCIS codes: 270.1670, 270.5585,A robust photonic quantum information network requires a means of storing and retrieving light on demand. For ensemble systems, storage schemes using Electromagnetically Induced Transparency (EIT) [1][2][3][4][5] and Raman coupling [6][7][8] show great potential. Another way of storing quantum information relies on photon echoes [9, 10] generated using controlled reversible inhomogeneous broadening (CRIB) [11][12][13][14]. It was recently shown that with a linear atomic frequency gradient introduced along the axis of light propagation, it is possible to obtain an efficient photon echo in the forward direction by simply reversing the sign of the gradient [15]. Experimental realisations of this Gradient Echo Memory (GEM) were performed using two-level atoms in a solid state system [15]. An advantage of GEM, and CRIB in general, is that a-priori knowledge of the temporal shape of the input pulse is not required for optimum storage [4,16]. In this paper, we propose a three-level GEM scheme, using far off-resonance Raman coupling to ground states. We also present a proof of principle experiment in warm rubidium vapour using a spatially varying Zeeman shift.We begin by demonstrating that Raman-coupled ground states can play an equivalent role to the twolevel atom considered in previous GEM work [15] in the far detuned and adiabatic limits. A three-level system is depicted Fig. 1(a) with a one-photon detuning ∆, a two-photon detuning δ(z, t) that can be varied in time and be made linear in space, a classical control beam Ω c and a weak quantum fieldÊ that we wish to store. The interaction Hamiltonian of the three-level system iŝ H =h(∆σ 33 + δ(z, t)σ 22 + gÊ †σ 13 + Ω * cσ 23 + h.c) (1) whereσ ii refers to the atomic population in the state |i , σ ij is the atomic coherence of the transition |i → |j and g is the atom-light coupling strength for the 1↔3 transition. Assuming that all the population is in the ground state |1 initially, and that the probe is weak (σ 11 ≃ 1), from the Heisenberg-Langevin equations, in a moving frame at the speed of light, we finḋwhere N is the linear atomic density and δ(z, t) = η(t)z, which is linear in z. The Langevin operatorsF 13 and F 12 account for noise coming from spontaneous emission (γ) and ground state decoherence (γ 0 ) respectively. It has ...

The micro-Wilhelmy method is a well-established method of determining surface tension by measuring the force of withdrawing a tens of microns to millimeters in diameter cylindrical wire or fiber from a liquid. A comparison of insertion force to retraction force can also be used to determine the contact angle with the fiber. Given the limited availability of atomic force microscope (AFM) probes that have long constant diameter tips, force-distance (F-D) curves using probes with standard tapered tips have been difficult to relate to surface tension. In this report, constant diameter metal alloy nanowires (referred to as "nanoneedles") between 7.2 and 67 microm in length and 108 and 1006 nm in diameter were grown on AFM probes. F-D and Q damping AFM measurements of wetting and drag forces made with the probes were compared against standard macroscopic models of these forces on slender cylinders to estimate surface tension, contact angle, meniscus height, evaporation rate, and viscosity. The surface tensions for several low molecular weight liquids that were measured with these probes were between -4.2% and +8.3% of standard reported values. Also, the F-D curves show well-defined stair-step events on insertion and retraction from partial wetting liquids, compared to the continuously growing attractive force of standard tapered AFM probe tips. In the AFM used, the stair-step feature in F-D curves was repeatably monitored for at least 0.5 h (depending on the volatility of the liquid), and this feature was then used to evaluate evaporation rates (as low as 0.30 nm/s) through changes in the surface height of the liquid. A nanoneedle with a step change in diameter at a known distance from its end produced two steps in the F-D curve from which the meniscus height was determined. The step features enable meniscus height to be determined from distance between the steps, as an alternative to calculating the height corresponding to the AFM measured values of surface tension and contact angle. All but one of the eight measurements agreed to within 13%. The constant diameter of the nanoneedle also is used to relate viscous damping of the vibrating cantilever to a macroscopic model of Stokes drag on a long cylinder. Expected increases in drag force with insertion depth and viscosity are observed for several glycerol-water solutions. However, an additional damping term (associated with drag of the meniscus on the sidewalls of the nanoneedle) limits the sensitivity of the measurement of drag force for low-viscosity solutions, while low values of Q limit the sensitivity for high-viscosity solutions. Overall, reasonable correspondence is found between the macroscopic models and the measurements with the nanoneedle-tipped probes. Tighter environmental control of the AFM and treatments of needles to give them more ideal surfaces are expected to improve repeatability and make more evident subtle features that currently appear to be present on the F-D and Q damping curves.

Quantum memories are an integral component of quantum repeaters-devices that will allow the extension of quantum key distribution to communication ranges beyond that permissible by passive transmission. A quantum memory for this application needs to be highly efficient and have coherence times approaching a millisecond. Here we report on work towards this goal, with the development of a 87 Rb magneto-optical trap with a peak optical depth of 1000 for the D2 F = 2 → F = 3 transition using spatial and temporal dark spots. With this purpose-built cold atomic ensemble we implemented the gradient echo memory (GEM) scheme on the D1 line. Our data shows a memory efficiency of 80 ± 2% and coherence times up to 195 µs, which is a factor of four greater than previous GEM experiments implemented in warm vapour cells.

Deterministic optical quantum logic requires a nonlinear quantum process that alters the phase of a quantum optical state by π through interaction with only one photon. Here, we demonstrate a large conditional cross-phase modulation between a signal field, stored inside an atomic quantum memory, and a control photon that traverses a high-finesse optical cavity containing the atomic memory. This approach avoids fundamental limitations associated with multimode effects for traveling optical photons. We measure a conditional cross-phase shift of π=6 (and up to π=3 by postselection on photons that remain in the system longer than average) between the retrieved signal and control photons, and confirm deterministic entanglement between the signal and control modes by extracting a positive concurrence. By upgrading to a state-of-the-art cavity, our system can reach a coherent phase shift of π at low loss, enabling deterministic and universal photonic quantum logic.cross-phase modulation | photonic quantum gate | cavity quantum electrodynamics | electromagnetically induced transparency | single-photon Kerr nonlinearity U niversal quantum gates (1, 2) can be implemented with an interaction that produces a conditional π-phase shift by one qubit on another (3). For photonic qubits, this requires an as-ofyet-unrealized strong cross-phase nonlinear interaction at the single-photon level. Photons do not directly interact with each other, and hence must be interfaced in a medium with a giant nonlinearity while preserving optical coherence (4, 5). The strong nonlinearities introduced by interacting Rydberg atoms (6-9) and cavity quantum electrodynamic (cQED) systems (10-12) have led to the observation of up to π-phase shifts between two propagating photons in the same mode. This type of quantum phase switch can be used to sort photons and implement a Bell state analyzer (13). The realization of a deterministic and universal optical gate, however, requires cross-phase modulation between distinct optical modes. Using a photon-atom gate in a cQED system, photon-photon entanglement (14) has been demonstrated. However, large conditional photon-photon phase shift remains an experimental challenge. For light pulses propagating in nonlinear fibers (15) and nonlinear slow-light media (16, 17), cross-phase modulation on the order of microradians per photon has been observed. In a pioneering cQED experiment two decades ago, Turchette et al. (18) measured the average polarization rotation of a weak continuous probe beam by another beam copropagating in the same cavity, and extrapolated a nonlinear phase shift of 0.28 rad per photon. However, the characteristic time of the nonlinearity (the cavity lifetime) in that experiment was much shorter than the photon wavepacket duration necessary to spectrally separate the two modes, which precludes the modulation of the entire wavepacket (19). Very recently, a much smaller but conditional cross-phase modulation of 18 μrad by a single postselected photon was measured in a nonlinear slow-light syste...

We demonstrate the feasibility of levitating a small mirror using only radiation pressure. In our scheme, the mirror is supported by a tripod where each leg of the tripod is a Fabry-Perot cavity. The macroscopic state of the mirror is coherently coupled to the supporting cavity modes allowing coherent interrogation and manipulation of the mirror motion. The proposed scheme is an extreme example of the optical spring, where a mechanical oscillator is isolated from the environment and its mechanical frequency and macroscopic state can be manipulated solely through optical fields. We model the stability of the system and find a three-dimensional lattice of trapping points where cavity resonances allow for build up of optical field sufficient to support the weight of the mirror. Our scheme offers a unique platform for studying quantum and classical optomechanics and can potentially be used for precision gravitational field sensing and quantum state generation.Recently much effort has been directed toward the development of new fabrication methods and experimental techniques for controlling optomechanical interactions at the quantum level [1,2]. Optomechanical effects have been observed in mechanical objects with masses ranging from femtograms, as in nano-optomechanical systems [3], to kilograms in the case of gravitational wave antennae [4]. Reaching the quantum regime in optomechanical systems is fundamentally interesting as one is then in a position to prepare macroscopic quantum states, which can, for example, be employed in tests of large-scale quantum decoherence [5] and models of gravity [6,7]. The main barrier to reaching the quantum regime is thermalization resulting from intrinsic coupling to environmental reservoirs. This is generally hard to avoid since most mechanical oscillators are supported by some mechanical structure that acts as a bridge for thermal fluctuations. One method to limit thermalization is to operate in cryogenic environments. Nevertheless, the dissipation of energy through the mechanical support still contributes significantly to the decoherence of the mechanical state [8]. Fabrication of a phononic-band gap structure into the substrate [9] has been proposed as one way to reduce the dissipation. Optical trapping [10] and levitation [11][12][13][14] have also been suggested as possible routes to low-dissipation quantum optomechanics. In the recent proposals, despite the mechanical support being completely removed, scattering from the levitated object leads to interaction with the environment and lowering of optomechanical coupling.Radiation pressure within an optical resonator can be used to couple the mechanical oscillations of a suspended cavity mirror with the optical mode [15,16]. Such coupling between optical and mechanical systems can be used for a variety of applications, including precision measurement [17], the creation of a mechanical quantum harmonic oscillator [15,[18][19][20], control of quantum macroscopic coherence [21], the generation of squeezed light for quantum informa...