The molecular microscope will enable us to detect individual molecules in a 2D square optical lattice with single-site resolution. This powerful device will be used for the direct detection of the novel quantum phases predicted for polar molecules in optical lattices. This ambitious project has the potential to transform our ability to control and study ultracold molecules in optical lattices. However, there are many technical challenges to overcome. Below, we discuss the implementation of the microscope using molecules created by atom association, as this approach currently looks the most promising for loading the lattice. We will focus on RbCs, though the techniques could equally well be employed with KCs and CsYb molecules.

Atomic Quantum Gas Microscopes

The atomic quantum gas microscope consists of a high-resolution optical imaging system integrated with a 2D optical lattice confining the atoms. The spacing between lattice sites is typically ∼500 nm. Resolving individual sites requires a custom microscope objective lens with a very high numerical aperture situated close to the plane of the lattice, as shown below.

Typical experimental apparatus for an atomic quantum gas microscope. A combination of a microscope objective and solid immersion lens close to the atoms typically achieves an effective numerical aperture of greater than ∼0.8. The atoms are illuminated from the side by molasses/Raman light, and the scattered fluorescence light is collected by the objective lens and projected onto a CCD camera.

The Molecular Microscope

We will need to modify the techniques developed for atoms in order to work with molecules.

Working distance and lattice spacing: Atomic gas microscopes typically position the 2D plane of atoms close to the surface of a solid immersion lens. The stray charge that can build up on dielectric surfaces can cause significant Stark shifts of the transitions used for STIRAP. Hence, in a molecule experiment it is important to limit the proximity of such surfaces where charge build-up can occur. In addition, the imaging apparatus must cater for electrodes positioned around the molecules in order to produce lab-frame dipole moments. The optics for a molecular microscope should therefore have a longer working distance than current atom microscopes. This reduces the numerical aperture of the imaging system, which ultimately reduces the spatial resolution. Fortunately, the large dipole moments of molecules produce measurable interactions over a relatively long range.

Dissociating molecules: To detect molecules produced by association, we reverse the STIRAP and magnetoassociation steps to return to pairs of free atoms which can then be imaged using the closed transition used for laser cooling. The established detection protocol is inappropriate because both atoms would quickly be lost through light-assisted collisions. Our solution is to use a species-specific vertical “pinning” lattice during the imaging sequence.

Species-specific trapping: To circumvent the problem of imaging doubly occupied lattice sites, we will separate the two species into different vertical sites during the imaging sequence, through a careful choice of wavelength. The figure below illustrates how we will use the species-specific lattice potentials to implement the microscope imaging.

Scheme for separating Rb and Cs atoms for imaging. Coloured arrows indicate direction of lattice beams. (a) We begin by loading a 2D square lattice with molecules. (b) Following the evolution in the lattice, the molecules are dissociated back into atoms. We choose a trapping wavelength where both Rb and Cs seek the high intensity regions, and hence share lattice sites. (c) By switching the lattice light in the vertical (z) direction to 830 nm the Rb and Cs are spatially separated for imaging.

Lattice loading: A key challenge facing the whole field is to load molecules into optical lattices with a high filling factor. The crucial step is to produce a Mott-insulator state with one atom of each species on each lattice site. This is difficult because of many competing differences between the two atomic species, e.g., different lattice potentials lead to different tunneling rates, different scattering lengths change the on-site interactions, the interspecies scattering length can lead to immiscibility, different density distributions are often not matched to one another. We will use our existing RbCs experiment to refine loading techniques, in combination with theoretical modelling of the lattice filling. An advantage of the molecular microscope is that defects in the lattice filling can be located and then accounted for in the analysis and modelling.

Proposed Apparatus

The development of a molecular microscope requires the construction of a new dedicated apparatus. Our existing RbCs apparatus does not have the optical access to add high-resolution imaging optics. We will draw on the established methods employed in atomic quantum gas microscopes and build an apparatus with a multi-chamber design where the initial magneto-optical traps and early-stage cooling are carried out in a separate chamber from the lattice, high-resolution imaging and molecule production.