We present a theoretical model for formation of molecules in an optical lattice well where a resonant coupling of atomic and molecular states is provided by small oscillations of a magnetic field in the vicinity of a Feshbach resonance. As opposed to an adiabatic sweep over the full resonance, this provides a coherent coupling with a frequency that can be tuned to meet resonance conditions in the system. The effective Rabi frequencies for this coupling are calculated and simulations show perfect Rabi oscillations. Robust production of molecules with an adiabatic sweep of the modulation frequency is demonstrated. For very large oscillation amplitudes, the Rabi oscillations are distorted but still effective and fast association is possible.PACS numbers: 03.75. Nt, 03.75.Ss, 36.90.+f Creation of molecules from ultracold atoms has been successfully achieved by making an adiabatic sweep of the magnetic field across a Feshbach resonance [1,2,3,4,5,6]. Feshbach resonances can be used to create molecules, because of the transformation of a scattering state into a new bound state, when the scattering length changes sign. This occurs when the magnetic field varies near a Feshbach resonance [7]:Here a bg is the background value of the scattering length, and B 0 and ∆ are the location and width of the resonance. In a recent publication [8] it was demonstrated that near Feshbach resonances, molecules can be created by harmonic modulation of the magnetic field:where ω B is resonant with the energy difference between the atomic and the molecular state at the magnetic field B ′ . It is the purpose of this article to analyze this process for two atoms in an optical lattice well, which is a very clean system for which many-body effects can be neglected. One argument in favour of producing the molecules in the lattice resonantly rather than by sweeping across the Feshbach resonance is that it enables the preparation of well-controlled coherent superposition states of atomic and molecular components. Ultracold molecules in optical lattices have been created by photoassociation [9, 10], and proposals for creating heteronuclear molecules have also been presented [11]. In [10], Rabi oscillations have been seen with this method. In [12,13] the production of molecules via Feshbach resonances in a 1D confining lattice potential has been studied and very recently the first successful creation of molecules in separate wells of a 3-dimensional optical lattice by sweeping across a Feshbach resonance has been reported [14]. As it turns out these molecules can be very long-lived (up to 700 ms) since they do not suffer from decay due to inelastic collisions. -0.6 -0.4 -0.2 0 0.2 0.4 0.6 -15 -10 -5 0 5 10 (B-B 0 )/∆ Energy (in units of ) hw Molecular state a (in units of 1000 a ) sc 0(1)(3) Figure 1: The s-wave spectrum of the potential (3) around the B0 = 155 G Feshbach resonance in 85 Rb in a harmonic oscillator trap with ω = 2π· 30 kHz. The variation of the scattering length asc as a function of B is shown as a dot-dashed line. Note that th...

We describe the collisional interaction between two different atoms that are trapped in a harmonic potential. The atoms are exposed to a magnetic field, which is modulated in the vicinity of an s-wave Feshbach resonance, and we study the formation of molecular bound states and excited states of the trapped system with non-trivial angular correlations.

We demonstrate a method to make mixtures of ultracold atoms that does not make use of a two-species magneto-optical trap. We prepare two clouds of 87Rb atoms in distinct magnetic quadrupole traps and mix the two clouds by merging the traps. For correctly chosen parameters the mixing can be done essentially without loss of atoms and with only minor heating. The basic features of the process can be accounted for by a classical simulation of particle trajectories. Such calculations indicate that mixing of different mass species is also feasible, opening the way for using the method as a starting point for making quantum gas mixtures.Comment: 12 pages, 13 figures. Fig. 10 corrected. Fig. 13 updated with more points and better statistics. A couple of paragraphs rephrased and typos corrected. References update