Many imaging techniques, particularly in environmental transmission electron microscopy (ETEM), generate images with degraded signal-to-noise ratio, contrast and spatio-temporal resolution, which hamper quantification and reliable interpretation of data. Moreover, the extraction of structural information from these images relies on manual acquisition and local structural identification which does not allow statistical analysis of the data and necessarily introduces a human bias carried out at the post-processing stage. The general aim of the ARTEMIA project is to develop a ground-breaking deep learning-based framework for in situ microscopy in liquid and gaseous media allowing the automated, high throughput, real-time acquisition and analysis of ETEM image sequences.Our framework will integrate aberration-corrected in situ ETEM imaging using windowed liquid/gas nanoreactors with denoising and resolution enhancement scheme set up using convolutional neural network (CNN). For model training, datasets consisting of simulated liquid- and gas-phase TEM images will be generated by by atomistic simulations including instrumental noise and imperfections of the microscope optics. In the ARTEMIA project, the CNN models will be applied to the study of two crystalline samples with complementary structural characteristics and electron beam sensitivity, model gold nanoparticles (Au NPs) and microporous zeolite, in reactive gas and/or liquid environments. Our scientific aim will be to gain further mechanistic understanding ofthe growth of model Au NPs in liquid phase and their reactivity in oxidizing and reducing gas environments on one hand and the steaming process of beam-sensitive zeolite on the other hand. The consortium comprises three academic partners (MPQ, LEM, IPCMS) and an EPIC partner (IFPEN) with complementary expertise in liquid and gas ETEM, data science and image processing with special focus on deep learning approaches, atomic modelling and TEM image simulation.

Ultracold atoms have emerged as unique tools to study strongly correlated quantum systems. 50 years ago, an intriguing prediction was made by Fulde, Ferrell, Larkin and Ovchinnikov (FFLO) for a superconductor in a magnetic field with imbalanced electron spin populations. They predicted the existence of a superfluid phase where the order parameter is inhomogeneous and oscillates spatially across the sample. In SPIFBOX, we aim at producing and studying this FFLO phase using spin-imbalanced Fermi gases in a box-shaped potential where the density of atoms is uniform. The experimental part of the project will be realized at ENS with two experimental setups using lithium isotopes. The first one is already operational and the flat bottom potential is realized in 3 dimensions by the repulsive mean field of a Bose-Einstein condensate of lithium 7 mixed with the spin-imbalanced Fermi superfluid in an harmonic trap. The second setup is a new generation experiment with much greater flexibility where the flat bottom potential is realized optically using a digital micro-mirror device (DMD). This new machine will enable us to search for the FFLO phase in reduced dimensions for which theoretical predictions and numerical simulations predict a much wider domain of stability for the FFLO state in the phase diagram. The construction of this setup has already started and the SPIFBOX funding will be used to bring it to completion. The theoretical part of the project will be conducted by Giuliano Orso from Paris Diderot University, a specialist of spin imbalanced Fermi systems, and one post-doc that we wish to recruit with SPIFBOX funding. The theory team will determine the optimal conditions for producing the FFLO phase. In particular, exact solutions exist when the fermions live in one dimension for which the existence of the FFLO phase is clearly established. They will also construct the phase diagram when the Fermi gas is mixed with a Bose gas and they will explore the possible stabilization of the FFLO phase in two and three dimensions by controlling the Fermi-Bose interaction strength. Numerical simulations using DMRG and Monte-Carlo methods will be compared to the ENS experimental observations. With the powerful tools of atomic physics, in SPIFBOX we gather together the best experimental conditions for the observation of the FFLO state: no orbital coupling, no disorder, low dimensional samples, and, most importantly, direct spin-resolved imaging of the associated spatial modulation of the atomic cloud. We hope that by solving one of the most outstanding quantum many-body problems, the outcome of SPIFBOX will stimulate new theoretical and experimental concepts at the interface with condensed matter systems. Ultimately this advanced understanding of quantum matter will help to design new materials with unprecedented properties.

Metasurfaces composed of dielectric nanoresonators with subwavelength thickness allow for strong control over the properties of light in transmission and reflection. In the last years, this control has been extended to classical parametric processes like sum-frequency generation, with the engineering of polarization and directionality of the generated light and the enhancement of the nonlinear conversion efficiency. These properties make nonlinear metasurfaces promising also for quantum technologies, especially for creating very thin, yet efficient and highly tailorable sources of photon pairs. This perspective is encouraged by recent observations of photon-pair spontaneous parametric generation in nonlinear metasurfaces. However, the true strength of metasurfaces, which stems from their open-system nature with access to numerous photonic degrees of freedom, makes the related modeling of a nonlinear quantum process very challenging. Yet, such a description is needed to control this broadband generation and therefore engineer the properties of the down-converted pairs towards specific quantum states. To date, this lack of a model has hindered both a fundamental insight in the operation of metasurface photon-pair sources and their development for quantum optical technologies. The goal of MEGAPHONE is therefore to establish a methodology for an accurate, but also computationally efficient and physically insightful description of photon-pair generation in dielectric metasurfaces, and use this methodology to create metasurfaces with tailored biphoton quantum states, and drastically improved generation efficiencies with respect to the state of the art. To this end, we will resort to quasinormal modes, which are a class of modal expansion suited for describing electromagnetic systems with a large amount of outward radiation or internal losses. MEGAPHONE will investigate pair generation from nonlinear metasurfaces in different operating regimes, with periodic and non-periodic arrangements of both non-interacting and interacting dielectric nanoresonators. Through a combination of analytical and numerical approaches, MEGAPHONE will both investigate the open-system effects that can potentially decrease the fidelity of the generated state with respect to the ideal targeted quantum state, and single out possible approaches for increasing it. Based on these theoretical grounds, the MEGAPHONE partners will fabricate and experimentally demonstrate concrete examples of metasurface photon-pair sources designed for generating polarization entangled states in the well-established AlGaAs technological platform of the French partner, and spatially correlated states in the well-established LN technological platform of the German partner. The results of this project will open the way for the broad development of metasurface photon-pair sources in different areas of photonic quantum technologies, from free-space quantum communication to quantum imaging and sensing.

Quantum information processing holds the promise to improve classical techniques taking advantage on the unique properties (e.g. entanglement) of quantum bits. Laser cooled trapped ions are an ideal system for the realization of such an idea. Miniaturized surface ion traps allow for scalability, but a technology fully compatible with the conventional CMOS fabrication is still needed. Here we propose to develop hybrid ion-traps (HIT), which will combine the use of a glass substrate (ideal for trapping) with a silicon interposer substrate (for laser beam steering and electronic connections). The envisioned embedding technology of the glass interposer into the Si substrate is such that HIT will be compatible with through-silicon-vias (TSV) and with mass production. HIT will allow us to implement photonic integrated circuits for ion addressing and readout. This will open the way to new trap designs (arrays and/or annular traps), contributing to large scale development of quantum computing.

The goal of our project is to study the spin polarization on single molecular clusters to suggest reliable nanospintronics devices based on the fundamental understanding of interface spin properties down to the single molecule level. To understand in detail the physics of such systems at the nanometer scale, we will proceed in two steps, using an original combination of complementary techniques. Firstly, we will perform structural and magnetic characterizations of homemade model ferromagnetic/organic systems. Secondly, we will measure the spin transport at the molecular scale by spin polarized scanning tunneling microscopy (SP-STM). This project is fundamental in nature with the expectation of understanding spin polarized transport through molecular nano-sized objects. The induced spin polarization of molecular layers is one of the current hot topics in spintronics which needs an international effort to be understood. Two routes can be explored, depending on the nature of the molecule, either purely organic or organo-metallic. In the first case, the spin polarization can arise from a different coupling between ‘molecular orbitals’ and electrons of majority and minority spin. The engineering of the spin polarization devices in this case requires a deep understanding of the coupling between a molecule and a continuum of states. In the second case (where the need of state-of-the-art chemical synthesis is mandatory and included in the project), the idea is to induce a spin state in the molecule through a metallic ion and to achieve a coupling between the ferromagnetic electrode and this spin state, what includes the physics of the Kondo resonance with spin polarized electrodes. Since the injection of spin currents into Alq3 (tris-(8-hydroxyquinoline)Al) systems starts to be well understood, we will begin our investigation of this part using similar molecular systems that are spin active like Crq3 and Feq3. Furthermore, we will explore molecular systems that can switch between different spin states by external excitations. In both cases, it is of major importance to understand from the study of model systems the typical mechanisms that can lead to the highest spin polarization in the molecular layer. This can be achieved only by a strong interaction between theoreticians and experimentalists, chemists and physicists, which is the goal of our project. An important part of the project will be devoted to the modelling and theoretical investigation of the physical properties of molecular systems. Typically, calculations combining both ab initio and tight binding techniques will be developed in order to better understand the role of the different parameters (coupling to the electrodes, hybridizations, spin-orbit coupling, many body effects…) on the spin transport.