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.

This proposal addresses the two major roadblocks in the development of graphene for high-performance nano-optoelectronics, namely how to efficiently and reliably integrate them in pristine conditions in electronic devices, and how harness the exceptional properties of graphene. Specifically, proof of principle of ultra-thin body tunnel field effect transistors (UB-TFET) are proposed consisting of two-dimensional (2D) all epitaxial graphene/boron nitride heterostructures with a viable large scale integration scheme. Tunnel transistors are an efficient alternative to standard field effect transistors designs that are inefficient for graphene because of the lack of a bandgap. Importantly UB-TFET should overcome the thermal limitation of thermioic sub-threshold swing in common transistors. The TFET will be based on epitaxial graphene on SiC (epigraphene, or EG)/BN structures; the most advanced implementation will utilize the recently discovered exceptional conductance properties of epigraphene nano-ribbons that are quantized single channel ballistic conductors at room temperature. But having excellent graphene is far from having a device and the active component has to be integrated. This project is based on the fundamental realization that only (hetero)-epitaxial growth can provide the required atomic control for reliable devices. Epitaxial growth insures clean interfaces and precise orientation of the stacked layers, avoiding trapped molecules and the randomness inherent to layer transfer. However, despite this absolute requirement, very little progress has been made up to now to grow large 2D dielectric on graphene; most dielectric deposition needs chemical modification of the graphene surface for adhesion, which invariably compromises the graphene electronic performance. Hexagonal boron nitride (h-BN) layers is considered the best substrate for graphene, but only micron size BN flakes are available, making the integration tedious, unreliable and impossible at large scale. In this proposal we will grow h-BN epitaxialy on epigraphene by metalorganic vapor phase epitaxy (MOVPE). As demonstrated in preliminary work by this three-team partnership, this technique provides exceptional unmatched graphene/h-BN epitaxial interfaces as required for high performance electronics, and immediate upscaling capabilities. The SiC/EG/h-BN heterostructure will give access to graphene properties in an exceptionally reproducible and clean environment, not otherwise available. Growth conditions will be investigated to produce ultra thin h-BN on epigraphene, which have not been achieved yet. This proposal will then follow two tracks to build UB-TFETs, demonstrating proof of principle of vertical and lateral BN/EG-based FETs. Our ultimate goal is to combine ballistic epigraphene nanoribbons in tunneling devices to enable a new generation of electronic devices. This is an extremely promising alternative to the standard FET paradigm that can enable ultra-high frequency operation as well as low power operation. This project is a tight well-focused partnership between three teams with a history of highly successful collaboration and perfect complementarity: CNRS-Institut Néel (Grenoble), CNRS/ONERA-Laboratoire d’Etude des Matériaux (Châtillon), and CNRS/Georgia Institute of Technology -UMI 2958 (Metz, in collaboration with GT Atlanta). We will build up on the important milestone of epitaxial h-BN growth on EG, towards critical development including ultra-thin BN and fabrication of tunnel transistors devices. IN will be in charge of providing epigraphene, will design and realized transistor devices and perform transport measurements; the UMI team will produce the BN epitaxial film and provide basic structural study for rapid optimization of the growth process; LEM will perform advanced structural and optical studies, in particular HR-TEM studies, critical to the layer characterization of ultra thin 2D films.

This project focuses on three iron-base alloys that have growing potential for high-temperature, high-strength and strong- magnet applications: Fe-Cr, Fe-Mn and Fe-Co. Because of the key role of magnetism, an innovative materials design based on advanced modeling approaches is necessary to control key properties of these materials. Such a design strategy requires the combination of (i) highly accurate methods to determine atomic features with (ii) efficient coarse-graining techniques to access target physical properties and to perform the screening of materials compositions. For the former, density functional theory (DFT) has for many materials classes already proven to be a highly successful tool. For Fe-based alloys, however, a critical bottleneck is the role that magnetic ordering, excitations and transitions have on thermodynamic, defect and kinetic properties. Therefore, a complete and accurate modeling of magnetism is urgently needed to address the materials-design challenges: the resistance to radiation damage related to the chemical decomposition in Fe-Cr, the grain-boundary embrittlement in ferritic Fe-Mn and the high-strength of austenitic Fe-Mn, and the phase ordering and the relative stability of a and ? phases in Fe-Co cannot be fully understood without properly accounting for the magnetic effects. The novelty of the current approach is twofold: First, on the DFT-side, we will make use of the recent important progress in treating magnetism in pure idealized Fe lattices, in order to go towards an accurate modeling of magnetic multi-component systems with point/extended defects, and beyond the standard collinear approximation. Second, we will develop new methods that allow us to bridge the gap between (i) highly accurate electronic calculations and (ii) large-scale atomistic thermodynamic and kinetic simulations for iron based alloys by – and this is decisive – fully taking into account the impact of magnetism on defect properties, diffusion and microstructural evolution. For the latter, lattice-based effective interaction models (EIMs) and tight-binding (TB) models will be developed based on data from DFT, including magnetic configurations, excitations and transitions. This will allow us to provide a coherent description of the role of magnetism on various properties of Fe-based alloys at different length scales and at finite temperature. It will further give us the ability to perform the optimization of key parameters controlling the relevant properties like phase decomposition in Fe-Cr, phase ordering in Fe-Co or decohesion of grain boundaries in Fe-Mn. Dedicated experiments in bulk alloys and along intergranular / interphase boundaries grown on demand will be performed in the project, which are essential for verifying the robustness of the theoretical predictions. The three chosen alloys exhibit a large variety of magnetic behavior. The methods developed and applied in this proposal are therefore expected to be transferrable to the modeling of other magnetic materials. The results of our simulations will lead to the improvement of thermo¬dynamic and diffusion databases and tools (such as DICTRA) that are nowadays routinely used in industrial R&D but that at present have difficulties in accounting for magnetism. In this way a better and more systematic understanding of the role of magnetism in Fe-based alloys will help to improve significantly the predictive power of the simulations and thus contribute to a more efficient and accurate development of new steel grades. Once fully implemented, the availability of such computational tools is expected to boost the efficiency, change the strategy in designing new steel grades and to form an important contribution for the future competitiveness of steelmakers.

EPOSBP project deals with black phosphorus (BP) which has joined the 2D materials family only very recently in 2014. The first representative of this new class of materials is Graphene, isolated ten years ago, which discovery sparkled an intense research activity. However, the lack of band gap in graphene has launched a quest for new 2D materials. The field has gradually been enriched by new contenders such as hexagonal Boron Nitride (hBN) and more recently the transition metal dichalcogenides family (e.g MoS2), leading to the emergence of a broad family of 2D van der Waals materials. In the specific optoelectronic field, the progress has remained limited due to the impossibility to gather together a direct bandgap AND a high carrier mobility within the same material. In this direction, black phosphorus (BP) has attracted an explosive interest since 2014 as it displays major properties for (opto-)electronic devices: (1) high hole and electron mobilities in thin layers of exfoliated BP (~3000 cm2 V-1 s-1). (2) high (~105) ON/OFF current ratio in a transistor configuration with ambipolar characteristics. (3) the BP bandgap is predicted to remain direct from the bulk to the monolayer, making it highly interesting for optical applications. The electronic structure near the Fermi level strongly depends on the number of layers, leading to a bandgap increase from the mid infrared (0.35 eV) in the bulk to the visible range (2 eV) in monolayers. Thus BP offers a unique spectral range in the 2D landscape. However, while early 2017 results seems to highlight BPs peculiar properties in ultrathin layers, only little is known from experimental measurements. Additionally, thanks to its natural low spin-orbit coupling, the BP could be expected to very efficiently preserve the spin lifetime of the carriers, as in graphene, but offering a semiconducting gap. This would be a unique opportunity for spin transport and spintronics. The objective of EPOSBP is to investigate these unique properties of BP and, capitalizing on them, to achieve new optically active flat materials from visible to mid-infrared. The dielectric response as well as the electronic behaviors and spin injection of BP transistor will be investigated to achieve tunable and electrically driven light emission. The project is decomposed into three tasks: 1) learning basics on BP properties aiming at defining a robust spectroscopic tools package for facilitating the integration of BP in devices, 2) fabrication and characterization of devices, 3) fabrication of BP transistors and observation of electroluminescence and efficient spin transport and 4) the demonstration of spin driven light emission. EPOSBP project constitutes a broad partnership that includes the best specialists and skills on 2D Black Phosphorus and Spintronics allied with specialists of most advanced relevant spectroscopic characterization techniques, integration of 2D materials in devices and specialists capable to demonstrate the potential of BP for innovative electroluminescence and tunable spin transport devices. The strong commitment of industrial partner Thales, with key interests in the semi-conductor area, is a strong enabler toward potential TRL rise of the project.

Industry is currently taking the lead to develop applications based on the genuine properties of Single Walled Carbon Nanotubes (SWNTs), that are their outstanding strength and aspect ratio, and ability to display metallic or semi-conducting characteristics, depending on their chiral structure. From a fundamental point of view, and also to make these applications commercially viable, finding a way to grow, on demand, metallic (m-) or semi-conducting (sc-) SWNTs with a reasonable yield and good selectivity, remains the biggest issue. The Catalytic Chemical Vapor Deposition (CVD) synthesis of SWNTs, that takes place at high temperature (600-1200 C), and in a complex chemical environment, is still not completely understood, though recent experiments, reporting a chiral selective growth, ignited a burst of questions, and new efforts to tackle this issue. In this context, the goal of the GIANT project is to build upon recent breakthrough results obtained in the understanding of the growth mechanisms, to gain an effective control of SWNT structure during their synthesis. In a previous project, involving the same partners, the importance of controlling the growth mode characterizing the geometry of the tube / catalyst nanoparticle (NP) interface during the growth, has been emphasized. A thermodynamic modeling relating interfacial energies to the resulting tube chirality has been developed. We also showed that using bi-metallic NPs, and fine tuning the growth conditions, led to a better SC/M selectivity. The underlying idea is now to focus on the chemistry and structure of the tube / NP interfaces using dedicated new experiments. Guided by the understanding brought about by our modeling, we will develop new catalysts, by different methods, including our original route based on the grafting and calcination of Prussian Blue Analogs, that enables to form dispersed, stable bimetallic and carbide NPs with controlled stoichiometry. Real time, in situ investigations will shed new light on observations that were previously done after growth. A strong asset of this project will be the use of the NanoMAX HR-TEM facility, that combines state of the art environmental Transmission Electron Microscopy (TEM), with original developments of the gas injection system, that make it perform under the same conditions as the UHV-CVD setup used in the laboratory. The chemical and structural evolution of the catalysts will be also investigated in situ and real time in a dedicated facility (FENIX) at LPICM. Systematic cross-checking between TEM (imaging and diffraction) and Raman assignments of the chiral distributions of produced tubes, and a comparison of in situ data with advanced post growth characterizations will be performed. Theory and modeling will be in constant interaction with experiments, either to guide the choice and modifications of the catalysts or to help the interpretation of the results. The thermodynamic modeling of the interface will be extended to include growth kinetics, and detailed atomistic Monte Carlo computer simulations of two kind of systems with different affinities for C (from NiPt to W or Mo carbide NPs) will be performed, to check their influence on the growth modes and resulting chiralities. Each catalytic system family (NiPt, CoMo and CoW, WC or Mo2C, carbon precursors and growth conditions) will be iteratively analyzed, tested for its selective growth ability, and, if successful, eventually used for producing m- or sc-SWNTs incorporated in different types of devices (sensors, transistors, field emission tips …). A successful outcome of the project will be the identification of selective catalytic systems offering either sc- or m-selectivity, with a reasonable yield, and the understanding of the underlying mechanisms.