To be updated soon
Current research
The long term goal of all experiments is to develop new methods to manipulate particles in a regime where the quantum mechanical aspects dominate their behavior and their properties. On the one hand, this should lead to new tools that allow one to probe physical laws and to measure fundamental constants with increasing precision. On the other hand, the progress of experimental methods also drives the advances in our understanding of the ever mysterious, beautiful, accurate, yet deeply dissatisfying structure of quantum mechanics. This interplay between theoretical concepts and experimental realizations promises to be very fertile in fields such as quantum control, quantum feedback and its limits, many-particle quantum systems, and many-particle entanglement (quantum computing). In particular, we are developing experimental methods to extend our control within the many-body Hilbert space.
Coupling many atoms to single photons: single-photon source and beyond
The collective interaction of many atoms with a single mode of an optical resonator also allows one to prepare entangled states of large samples of atoms by detection of the photons emitted by the atoms. When the atoms interact with the mode in a way that makes it fundamentally impossible to determine which atoms, e.g. emitted a photon into the resonator, the system must be described by an entangled state of the atomic ensemble (Dicke state). Under appropriate conditions, these states interact collectively, and therefore very strongly with photons.
Following a slightly modified version of the quantum repeater proposal by Duan, Lukin, Cirac, and Zoller [1], we have built a system where a sample of atoms is made to conditionally generate a single photon on demand following the (random) detection of a previous single photon. Currently we are able to store the excitation corresponding to a single photon for 2 us as a ground-state polarization grating (spin wave) in the atomic sample containing 106 atoms, and to convert this excitation into a single photon with an efficiency near 90%. The storage time is limited by the Doppler effect, i.e. by the random thermal motion of the atoms that destroys the holographic grating. We are working on increasing the storage time into the range of milliseconds, and possibly seconds, by decreasing the Doppler effect, and ultimately removing it altogether by strong confinement of the atoms (Lamb-Dicke regime). We also expect to be able to further increase the read efficiency, producing an even better approximation to a single-photon Fock state on demand.
[1] L.M. Duan. M.D. Lukin, J.I. Cirac, and P.Zoller, Nature 414, 413 (2001).
New laser cooling methods for atoms, ions or molecules: Cavity Doppler cooling and cavity sideband cooling by coherent scattering
Laser cooling of atoms has not only supplied the basis for the control and manipulation of matter at the quantum limit, e.g. in form of Bose-Einstein condensation, but has also resulted in a number of important applications and devices, many of which are tied to precision measurements and atomic clocks. However, laser cooling has so far been limited to atoms with a particular internal structure, and the cooling of molecules or even of atoms with a complicated level scheme has not been possible. If we could learn how to cool, trap and manipulate larger molecules in the same way as atoms, this would open the door for important developments in chemistry and possibly even in biology.
Doppler cooling [2] is the dominant laser cooling mechanism at all but the lowest temperatures. In Doppler cooling counterpropagating laser beams are tuned to the red of a closed transition between an atomic ground and an atomic excited state. An atom that is moving towards a laser beam will experience photons that are blue-shifted into resonance with the atomic transition, while photons from a beam propagating in the same direction as the atom will be red-shifted further out of resonance. The momentum transfer associated with the preferred absorption of photons from the counterpropagating beam leads to slowing and cooling of the atoms, while the randomly emitted photons on average do not contribute to the force. The net effect is cooling to temperatures in the millikelvin to microkelvin range, corresponding to atomic velocities in the range of a few millimeters to a few centimeters per second.
This principle behind laser cooling also entails its limitation. Since the momentum “kick” associated with each photon absorption event is much smaller than the momentum of a thermal atom, a larger number of absorption-emission events (on the order of thousand or more) is required to significantly change the atom’s velocity. Therefore laser cooling has only been demonstrated with atoms that can be optically cycled many times back to their initial ground state. However, most atoms (and all molecules) have multiple ground states to which the excited state can decay. Once the atom reaches a different ground state, the laser no longer has the correct detuning relative to the atomic transition, and the cooling stops. In particular, molecules have many vibrational and rotational levels, and consequently no laser cooling of molecules has been demonstrated.
The novel proposed method [3,4] is based on coherent scattering, rather than on spontaneous emission from an excited state. Coherent scattering dominates when the laser is far detuned from atomic or molecular transitions and when its intensity is not too large. Coherent scattering is generic to all polarizable particles, independent of their internal structure, and describes the emission of radiation by an oscillating atomic dipole that is driven by an external (classical) electric field. The basic idea behind the proposed technique is that energy is conserved in the scattering process, and that therefore events where the scattered photon carries away a larger energy than the incident-photon energy are accompanied by a corresponding reduction of the atom’s energy. Such scattering events can be enhanced in an optical resonator that is tuned to be resonant with a frequency that is higher than that of the incident light. The new cooling mechanism depends only on the finesse (i.e. on the reflectivity of the cavity mirrors) and on the detuning of the photons relative to the cavity resonance, while it is independent of the detuning relative to atomic transitions. Therefore this new technique should be generic and be applicable to any sort of material. The target can be an atom in different ground states, a molecule in different rotational and vibrational states, or possibly even a scattering center (impurity) inside a solid. The only requirement is that at the given intensity and laser frequency the emission rate by the scatterer is large enough to produce efficient cooling. The cooling power is given by the product of the scattering rate and the energy difference between the incident and the scattered photon [3].
When we performed what was to be a proof-of-principle experiment with cesium atoms inside an optical resonator, we found very much to our surprise that both the light emitted by the atoms into the cavity, and the cavity-emission-induced forces on the atoms were much larger than expected for independent atoms [5,6]. In some cases [6] the observed emission and forces exceeded the predicted single-atom values [2-4] by a factor of a few thousand! The observed laser-like collective emission by the atomic sample required an optical gain mechanism. For not too large light-atom detunings, where the atomic multilevel structure is relevant, we obtained experimental evidence that the optical gain is Raman gain between differently populated magnetic sublevels. In the limit of large detuning, where the atomic multilevel structure is negligible, and the atoms behave like classical light scatterers, Domokos and Ritsch from the University of Innsbruck suggested that the collective emission could be explained by a collective process where the atoms self-organize into a density grating [8]. Then the classical (Rayleigh) emission by individual atoms, proportional to the number of atoms N, would be transformed into Bragg scattering by the density grating, scaling as N 2. We have verified those predictions experimentally by measuring the phase of the light emitted into the cavity, and furthermore observed a very strong cooling of the sample's center-of-mass motion [7].
[2] T.W. Hänsch and A.L. Schawlow, Opt. Commun. 13, 68 (1975).
[3] V. Vuletic and S. Chu, Phys. Rev. Lett. 84, 3787 (2000).
[4] P. Horak, G. Hechenblaikner, K.M. Gheri, H. Stecher, and H. Ritsch, Phys. Rev. Lett. 79, 4974 (1997).
[5] V. Vuletic, H. W. Chan, and A. T. Black, Phys. Rev. A 64, 033405 (2001).
[6] H.W. Chan, A.T. Black, and V. Vuletic, Phys. Rev. Lett. 90, 063003 (2003).
[7] A.T. Black, H.W. Chan, and V. Vuletic, Phys. Rev. Lett. 91, 203001 (2003).
[8] P. Domokos and H. Ritsch, Phys. Rev. Lett. 89, 253003 (2003).