Cs Quantum Gas Microscope

A quantum gas microscope for strongly interacting topological phases.

We have built a new Quantum Gas Microscope experiment with bosonic Caesium atoms at LMU to study topological many-body phases of matter. We will make use of the unique possibilities offered by high-resolution imaging techniques to investigate topological many-body phenomena and out-of-equilibrium dynamics in these lattices.

We are currently looking for PhD and Master students!!

Quantum simulation of topological many-body phases

Neutral atoms in optical lattices offer exceptional opportunities for exploring complex quantum many-body systems, both in and out of equilibrium, within a highly controlled environment. A natural class of models readily accessible in these systems are Hubbard models, where particle dynamics are governed by the interplay between tunneling and on-site interactions. This facilitates fundamental investigations of the nature of the phase diagram of paradigmatic models such as the Fermi-Hubbard model. In contrast, quantum simulation of strongly correlated topological phases of matter requires the development of novel engineering techniques, as the relevant microscopic interactions do not naturally arise in these systems.

a Schematic of the flux ladder; b Phase diagram

a Schematic of the flux ladder; b Phase diagram

In this context, Floquet engineering—also known as periodic driving—has emerged as a powerful method. Over the past decade, it has been successfully employed to realize a variety of lattice models with topological band structures, and numerous experimental techniques have been developed to probe their geometric properties. However, with few exceptions involving interacting two-particle systems, experiments have been carried out almost exclusively in the non-interacting regime. While these studies clearly demonstrate the potential of Floquet engineering, extending it to many-particle strongly-interacting systems remains a significant challenge. The main challenge in this regard is heating due to the periodic drive, which eventually results in an infinite-temperature state.

In a recent study, were able to leverage the new capabilities of our cesium quantum gas microscope in order to realize a large Floquet-engineered many-body flux ladder with about 24 hard-core bosons at half filling in a low-entropy state, whose temperature we benchmarked to be on the order of the tunnel coupling. By tuning the coupling ratio K/J we explore the phase diagram which exhibits a transition from a Meissner to a vortex regime. These results represent a significant advance in Floquet many-body physics and marks an exciting first step toward realizing large fractional Chern insulators with cold atoms, opening the door to studying topological order and exotic anyonic excitations with microscopic resolution and control [A. Impertro et al., arXiv:412.09481 (accepted in Nature Physics)].

Developing novel observables for quantum gas microscopes

We recently demonstrated a novel technique for state-preparation and read-out based on optical superlattice potentials that enable local manipulations on the level of isolated bonds in the lattice.

Schematic drawing of the projection of the many-body wave function onto isolated double wells using a superlattice potential and the corresponding representation of the double-well manipulations on a Bloch sphere.

Schematic drawing of the projection of the many-body wave function onto isolated double wells using a superlattice potential and the corresponding representation of the double-well manipulations on a Bloch sphere.

Quantum gas microscopes have revolutionized quantum simulations with ultracold atoms, allowing to measure local observables and snapshots of quantum states. However, measurements so far were mostly carried out in the occupation basis. In this work, we demonstrate how all kinetic operators, such as kinetic energy or current operators, can be measured and manipulated with single bond resolution. Beyond simple expectation values of these observables, the single-shot measurements allow to access full counting statistics and complex correlation functions. Our work paves the way for the implementation of efficient quantum state tomography and hybrid quantum computing protocols for itinerant particles on a lattice. In addition, we demonstrate how site-resolved programmable potentials enable a spatially-selective, parallel readout in different bases as well as the engineering of arbitrary initial states [A. Impertro et al. Phys. Rev. Lett. 133, 063401 (2024)] .

Viewpoint in Physics highlighting our article: Visualizing Atom Currents in Optical Lattices

Unsupervised machine learning for single-site reconstruction in quantum gas microscopes

We demonstrated a novel technique based on a convolutional autoencoder which enables us to reconstruct the lattice occupation with high fidelity even though our lattice spacing is more than two times smaller than our imaging resolution (383.5nm lattice spacing vs. 850nm Rayleigh resolution).

(left) Architecture of the neural network. (right) example reconstruction of a Mott insulator with ~ 2500 atoms.

(left) Architecture of the neural network. (right) example reconstruction of a Mott insulator with ~ 2500 atoms.

In quantum gas microscopy experiments, reconstructing the site-resolved lattice occupation with high fidelity is essential for the accurate extraction of physical observables. For short interatomic separations and limited signal-to-noise ratio, this task becomes increasingly challenging. Common methods rapidly decline in performance as the lattice spacing is decreased below half the imaging resolution. Here, we present a novel algorithm based on deep convolutional neural networks to reconstruct the site-resolved lattice occupation with high fidelity. The algorithm can be directly trained in an unsupervised fashion with experimental fluorescence images and allows for a fast reconstruction of large images containing several thousand lattice sites. We benchmark its performance using a quantum gas microscope with cesium atoms that utilizes short-spaced optical lattices with lattice constant 383.5nm and a typical Rayleigh resolution of 850nm. We obtain promising reconstruction fidelities ≳96% across all fillings based on a statistical analysis. We anticipate this algorithm to enable novel experiments with shorter lattice spacing, boost the readout fidelity and speed of lower-resolution imaging systems, and furthermore find application in related experiments such as trapped ions [A. Impertro et al., Commun. Phys. 6, 166 (2023)] .

PhD theses from our team: