We report on collective non-linear dynamics in an optical lattice formed inside a high finesse ring cavity in a so far unexplored regime, where the light shift per photon times the number of trapped atoms exceeds the cavity resonance linewidth. We observe bistability and self-induced squeezing oscillations resulting from the retro-action of the atoms upon the optical potential wells. We can well understand most of our observations within a simplified model assuming adiabaticity of the atomic motion. Non-adiabatic aspects of the atomic motion are reproduced by solving the complete system of coupled non-linear equations of motion.PACS numbers: 32.80. Pj, 42.50.Vk, 42.62.Fi, Single atoms interacting with a few photons inside a high finesse cavity are an extensively studied key system of quantum optics. Only recently the significance of the atomic motion in this model system has been recognized and a wealth of new physics was found including cooling and trapping single atoms by single photons [1,2]. More recently, it has been pointed out that optical cavities with sufficient finesse and large mode volumes could provide a new means to cool and trap even large atomic samples by coherent scattering, with the promise of extending laser cooling to a wider class of atomic species and possibly even molecules [3,4,5,6,7]. A so far unexplored regime of atom-cavity interactions arises for large atom samples trapped in cavity-enhanced far off-resonant light fields, if the collective coupling strength (i.e. the light shift per photon times the atom number) exceeds the cavity resonance linewidth. For sufficiently high finesse of the cavity, the otherwise tiny retro-action of the moving atoms upon the light field becomes a significant feature of this system. This leads to an inherently collective character of the atomic motion and a corresponding non-linear dynamics of the intra-cavity field. Such collective long-range interactions might allow to create controlled quantum entanglement [8], a perspective possibly useful for quantum computation with neutral atoms. Other intriguing perspectives are the implementation of improved cooling schemes for atomic species otherwise not accessible, for example, as indicated in ref. [7]. Sympathetic cooling without collisions might become possible for low particle densities via thermalization of atom samples trapped at distant locations inside the cavity.In this paper we explore the collective motion of atoms trapped inside an optical standing wave formed by two mutually counterpropagating travelling wave modes in a high-finesse ring resonator. A fast servo tightly locks the laser frequency in resonance with one of the modes. Within a narrow window around the case of exactly symmetric pumping of both cavity modes, a stable optical lattice is formed. Asymmetric pumping on the order of a few percent yields surprisingly complex dynamics. The most striking feature is a drop of the intra-cavity intensity of the unlocked mode by more than an order of magnitude, if the incoupled powers deviate by onl...

We report the confinement of large clouds of ultracold 85 Rb atoms in a standing-wave dipole trap formed by the two counterpropagating modes of a high-Q ring cavity. Studying the properties of this trap, we demonstrate loading of higher-order transverse cavity modes and excite recoil-induced resonances.

We demonstrate a novel optical dipole trap which is based on the enhancement of the optical power density of a Nd:YAG laser beam in a resonator. The trap is particularly suited for experiments with ultracold gases, as it combines a potential depth of order 1 mK with storage times of several tens of seconds. We study the interactions in a gas of fermionic lithium atoms in our trap and observe the influence of spin-changing collisions and off-resonant photon scattering. A key element in reaching long storage times is an ultra-low noise laser. The dependence of the storage time on laser noise is investigated.OCIS: 020.7010 020.2070Far-detuned optical dipole traps are rapidly becoming standard tools for atomic physics at ultralow temperatures [1]. They allow trapping of practically any atomic species, and even molecules. In the field of quantum gases they allow trapping of mixed-state and mixed-species ensembles. The coupling of the atoms to the light field, which is small for far detuned traps, can be strongly enhanced by means of an optical resonator. Indeed, in cavity quantum electrodynamics experiments, single atoms have been trapped by a light field that corresponds to a single photon [2]. Optical resonators have also been used to sensitively detect optical fields in a quantum non-demolition way using cold atoms [3], and even open up new possibilities for optical cooling of atoms and molecules [4]. In this letter we demonstrate a resonator-enhanced dipole trap (REDT) for experiments with ultracold gases. We take advantage of the resonant enhancement of the optical power density and the corresponding trap depth. To suppress photon scattering and reach storage times of several tens of seconds, the detuning of the trap light from the atomic resonances is large. At the same time the trap volume and potential depth are large to transfer many atoms into the trap. We expect the high optical power density reached in the REDT will be useful in many contexts, for instance for trapping earth-alkali atoms, buffer-gas cooled atoms or cold molecules.Our primary interest is in spin mixtures of fermionic 6 Li, as a promising candidate system for the formation of Cooper pairs in an atomic gas [5]. In particular, we aim to study Feshbach scattering resonances which have been predicted at experimentally accessible magnetic fields, which may provide a binding mechanism for the Cooper pairs [6].To sufficiently suppress photon scattering, the trap light must be detuned from the 670-nm D lines of Li by few hundred nm. In addition, to capture atoms from a magneto-optical trap (MOT), the optical trap must have a depth of the order of the temperature of Li in a MOT (∼ 1 mK), and a similar spatial extension. The optical power required to create such a trap without resonant enhancement exceeds 100 W. Our REDT only requires a 1.2 W Nd:YAG laser (λ = 1064nm), the power density of which is resonantly enhanced 130-fold in a 15 cm near-confocal resonator (See Fig. 1). The resonator mirrors are placed outside the vacuum, which facilitates adjus...

An optical lattice with rubidium atoms ( 85 Rb) is formed inside a ring resonator with a finesse of 1.8 × 10 5 and a large mode volume of 1.3 mm 3 . We typically trap several times 10 6 atoms at densities up to 10 12 cm −3 and temperatures between 25 and 125 µK. Despite of the narrow bandwidth (17.3 kHz) of the cavity, heating due to intra-cavity intensity fluctuations is kept at a low level, such that the time evolution of the temperature is determined by evaporative cooling.

We study the collective motion of atoms confined in an optical lattice operating inside a high finesse ring cavity. A simplified theoretical model for the dynamics of the system is developed upon the assumption of adiabaticity of the atomic motion. We show that in a regime where the light shift per photon times the number of atoms exceeds the line width of the cavity resonance, the otherwise tiny retro-action of the atoms upon the light field becomes a significant feature of the system, giving rise to dispersive optical bistability of the intra-cavity field. A solution of the complete set of classical equations of motion confirms these finding, however additional non-adiabatic phenomena are predicted, as for example self-induced radial breathing oscillations. We compare these results with experiments involving laser-cooled 85 Rb atoms trapped in an optical lattice inside a ring cavity with a finesse of 1.8 × 105 . Temperature measurements conducted for moderate values of the atom-cavity interaction demonstrate that intensity-noise induced heating is kept at a very low level, a prerequisite for our further experiments. When we operate at large values of the atomcavity interaction we observe bistability and breathing oscillations in excellent agreement with our theoretical predictions.