Research topics

Many-body cavity QED

A gas of cold atoms loaded into an optical resonator opens up for the realizion of numerous interesting models and the possibility to study all kinds of light-matter phenomen, such as optomechanics and multi-partite entanglement. The atom-field interaction can either be dispersive, where the atomic motional degrees of freedom are coupled to the field degrees of freedom, or resonant, where also the internal atomic degrees of freedom play and important role. An outcome of the cavity field is that it induces an effective (infinite range) interaction between the atoms, meaning that the physics of these models are often rather different from those where the interaction steems from atomic s-wave scattering.

Own activities

Phase diagram of cold bosonic atoms confined in an optical resonator. Shaded areas are Mott insulating phases, while the white region is the superfluid phase.

• Hubbard-like physics. Lattice models are constructed either with an external optical potential or by utilizing standing wave modes of the cavity. In the latter case, the lattice becomes dynamical which in certain limits can be assigned an effective non-local Hubbard type Hamitltonian ( Phys. Rev. Lett. 100, 050401 (2008) , arXiv). The interplay between short range s-wave atomic scattering and long range cavity induced interaction may give rise to very exotic models not similar to others found in different fields of physics.
• Dicke physics. For 'transverse pumping' of a cold bosonic gas, at a critical pumping amplitude the atoms self-organize and a large number of photons scatter from the pump into the cavity ( Phys. Rev. A 81, 043407 (2010), arXiv). In certain cases, this transition is described by an open Dicke phase transition.

Ultracold gases in optical lattices

Optical lattices are used in order to realize lattice models and to enter the strongly correlated regime with cold atoms. The flexibility in setting up the lattice configurations, the low temperature, the large control of the atoms, and the isolation these systems allow for a plethora of lattice models to be realized. Ground state physics can be explored and especially various types of matter. The systems are also well suited for studies of non-equilibrium physics, like for example follwing in situ the many-body evolution after a quantum quench.

Own activities

Phase diagram of the Heisenberg XYZ model in an external field. On the vertical axis is the external field and on the horizontal the z coupling strength.

• Orbital physics. The periodicity of the system implies a band-structure spectrum. In particular excited bands can become degenrate which means that the states of individual atoms living on these bands can be ascribed an orbital structure. This additional degree of freedom makes the physics of these atoms much more rich than for atoms residing solely on the lowest energy band. Recently we showed how this can be utilized in order to construct quantum simulators of quantum magnetism (Phys. Rev. Lett. 111, 205302 (2013) , arXiv).
• Disorder. The most astonishing result of disorder is the phenomenon of localization. There are many open questions or not well understood topics in this field, especially in interacting systems - so called many-body localization. How does decoherence affect the many-body localization effect, how is the eigenstate thermalizaion hypothesis affected by decoherence, what happens when the disorder breaks some symmetries?

Cavity/circuit QED

This is the field of light-matter interaction when a few degrees of freedom can be sorted out. The archetype situation is that of a qubit copled to a single boson mode described by the Jaynes-Cummings model (or in the strong coupling regime by the Rabi model). Early experiments studied fundamental questions in quantum theory like entanglement and the classical-limit. Today, the goals are to reach stronger qubit-light couplings as well as in a controlled manner increase the number of degrees of freedom, both boson and spin, to explore quantum-many body physics.

Own activities

The two "energy surfaces" (in a semi-classical picture) of the Jaynes-Cummings model. The conical intersection at the origin gives rise to many intriguing properties like non-vanishing Berry phases.

• Quantum simulators. I am particularly interested in situations where cavity/circuit QED physics meets other fields like molecular ( Phys. Rev. A 78, 033833 (2008), arXiv) or condensed matter physics ( Phys. Rev. A 81, 051803(R) (2010), arXiv). Also collective phenomena and how they are affected by decoherence are catching my attention ( Euro. Phys. Lett. 90, 54001 (2010), arXiv).
• Quantum dynamics. These systems are suited for in situ studies of quantum dynamics, both for closed and open systems. One interesting new direction is optomechanical cavity-based systems which takes quantum mechanics to the boarder line of the 'classical world'. Manifestations of quantum phenomena in this regime is thereby particularly interesting, like the Josephson effect ( Phys. Rev. A 84, 021804(R) (2011), arXiv) and entanglement ( Phys. Rev. A 85, 033805 (2012), arXiv). I have also been interested in more fundamental questions like the adiabatic limit and how topology affects certain observables ( Phys. Rev. Lett. 108, 033601 (2012), arXiv).

Synthetic gauge fields

In the realm of quantum simulators it is desirable to have mathematically analogues models describing charge particles in external magnetic fields. In condensed matter physics the most famous example is the Quantum Hall effect. In terms of the fractional quantum Hall effects we still face many open questions and hence it is a goal to realize this regime with cold atoms. Constructing synthetic gauge fields could also be a way to explore topological phases with artificial matter.

Own activities

Interplay between a transverse spin Hall motion (a) and a longitudinal Bloch oscillation motion (b). Here the system is a Rashba spin-orbit coupled atomic condensate in a tilted square optical lattice.

• Cavities. The Rabi model, or in the language of molecular physics the Exb Jahn-Teller model, is a most simple example of a system driven by a spin-orbit coupling term. By considering multi-mode cavities it is possible to extend the example above to construct Rashba spin-orbit couplings ( Phys. Rev. Lett. 103, 013602 (2009), arXiv) and thereby study non-Abelian outcomes of such couplings ( Phys. Rev. A 81, 051803(R) (2010), arXiv). Other related interests are synthetic gauge theories in Jaynes-Cummings-Hubbard models.
• Cold atoms. In cold atomic systems I have mainly been studying manifestations of the synthetic gauge fields in the dynamics of condensates, like in Berry phases ( Phys. Rev. A 79, 043627 (2009), arXiv) and the spin Hall effect ( Phys. Rev. A 82, 043620 (2010), arXiv). Currently I am also considering sort of 'dynamical' gauge fields.

Quantum chaos and quantum many-body dynamics

In a classical sense, chaos acnnot exist in quantum systems. Still, quantum systems where their corresponding classical models are chaotic show some generic features. Understanding these properties are of great relevance since it is today possible to isolate quantum systems and study their evolutions after a quench where system parameters are suddenly changed.

Own activities

Densities of an atomic condensate after a long time following a quantum quench. In the upper plot full quantum simulations, and in the lower corresponding semi-classical truncated Wigner simulations. The irregular structures indicates "quantum thermalization", and the enhanced densities are so called quantum scars.

• Quantum signatores in chaotic systems. One route to explain how chaos can appear is by introducing decoherence in your system, for example via a measurements. This approach has been successful in systems have a classical counterpart. Recently we showed that this need not be true when the system is lacking a classical limit (J. Phys. B. 46, 224015 (2013) ,arXiv). Chaos or non-integrability in the corrsponding classical systems has also been taken as a prerequisite for quantum thermalization. This cannot be true in general, take for example a disordered system supporting localization. I recently showed that the picture is in general much more complex, especially in systems lacking a propper classical limit (J. Phys. B. 46, 224016 (2013),arXiv). The Eigenstate thermalization hypotesis says that if the eigenstates of a given system shows a weak energy dependence, then the system will thermalize. In classical chaotic systems the dynamics can be "mixed" with both regular and chaotic evolution (In terms of Poincare maps, there can be 'islands' of regular dynamics in a chaotic 'sea'.). We have shown that such islands survive also in the quantum systems and thermalization occurs in the chaotic sea but not in the islands ( Phys. Rev. A 87, 013624 (2013), arXiv).

Open quantum systems and quantum measurements

A quantum system is open when it is coupled to some other quantum system, typically a reservoir or a measurement device. In optical systems, the coupling to reservoirs can often be assumed Markovian meaning that any information leaking from the system to the reservoir is for ever lost. The evolution of the system is no longer unitary and an initial pure state of the system may become mixed.

Own activities

This figures shows how pure decoherence "kills" fine quantum interference structures in the Wigner distribution: From (a) to (d) the decoherence strength is increased from fully coherent evolution in (a).

• Decoherence. The (markovian) coupling of a system to a thermal reservoir implies that the system thermalizes. For driven systems the situation is more complex, and the system can actually approach a pure state, a so called dark state that is prone to the decoherence and may be highly non-classical. Here the reservoir is an asset; the state preparation is forced by the reservoir. Related to this, I'm interested in the interplay between localization and decoherence. Another scenario where the dynamics can become highly non-trivial is in non-Markovian open systems ( Phys. Rev. A 83, 052103 (2011),arXiv).
• Measurement induced localization. Quantum measurements unavoidably affects the measured system. Nondemolition measurements characterizes measurements of an observable that commutes with the Hamiltonian, and repeated measurements will thereby give identical outcomes ( Phys. Rev. A 80, 053609 (2009),arXiv). In chaotic systems, a continuous position measurement can localize the phase space distribution and reproduce classical chaos (J. Phys. B. 46, 224015 (2013) ,arXiv).